Examenos

HF text 0.4 · Full score

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Public index

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Private same-skill component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Private new-domain component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Raw public index

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Public breadth index

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Equated private same-skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Equated private new domains

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Full contribution · public

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Published weighted contribution; not a distinct benchmark execution.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Full contribution · same_skills

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Published weighted contribution; not a distinct benchmark execution.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Full contribution · new_domains

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Published weighted contribution; not a distinct benchmark execution.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Knowledge · raw

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Count is contributing benchmarks, not requests. Source area context: {"id": "knowledge", "raw": 0.6554, "skill": 0.5714, "combined": 0.4779, "coverage": 1, "pending": 0, "n": 11}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 11

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Knowledge · skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Count is contributing benchmarks, not requests. Source area context: {"id": "knowledge", "raw": 0.6554, "skill": 0.5714, "combined": 0.4779, "coverage": 1, "pending": 0, "n": 11}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 11

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Knowledge · combined

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Count is contributing benchmarks, not requests. Source area context: {"id": "knowledge", "raw": 0.6554, "skill": 0.5714, "combined": 0.4779, "coverage": 1, "pending": 0, "n": 11}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 11

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Language · raw

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Count is contributing benchmarks, not requests. Source area context: {"id": "language", "raw": 0.8006, "skill": 0.7222, "combined": 0.6213, "coverage": 1, "pending": 0, "n": 9}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 9

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Language · skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Count is contributing benchmarks, not requests. Source area context: {"id": "language", "raw": 0.8006, "skill": 0.7222, "combined": 0.6213, "coverage": 1, "pending": 0, "n": 9}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 9

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Language · combined

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Count is contributing benchmarks, not requests. Source area context: {"id": "language", "raw": 0.8006, "skill": 0.7222, "combined": 0.6213, "coverage": 1, "pending": 0, "n": 9}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 9

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Retrieval · raw

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Count is contributing benchmarks, not requests. Source area context: {"id": "retrieval", "raw": 0.7709, "skill": 0.6883, "combined": 0.676, "coverage": 1, "pending": 0, "n": 6}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 6

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Retrieval · skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Count is contributing benchmarks, not requests. Source area context: {"id": "retrieval", "raw": 0.7709, "skill": 0.6883, "combined": 0.676, "coverage": 1, "pending": 0, "n": 6}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 6

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Retrieval · combined

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Count is contributing benchmarks, not requests. Source area context: {"id": "retrieval", "raw": 0.7709, "skill": 0.6883, "combined": 0.676, "coverage": 1, "pending": 0, "n": 6}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 6

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Tools · raw

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Count is contributing benchmarks, not requests. Source area context: {"id": "tools", "raw": 0.8378, "skill": 0.8186, "combined": 0.8038, "coverage": 1, "pending": 0, "n": 5}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 5

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Tools · skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Count is contributing benchmarks, not requests. Source area context: {"id": "tools", "raw": 0.8378, "skill": 0.8186, "combined": 0.8038, "coverage": 1, "pending": 0, "n": 5}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 5

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Tools · combined

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Count is contributing benchmarks, not requests. Source area context: {"id": "tools", "raw": 0.8378, "skill": 0.8186, "combined": 0.8038, "coverage": 1, "pending": 0, "n": 5}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 5

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Arts · raw

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Count is contributing benchmarks, not requests. Source area context: {"id": "arts", "raw": 0.7016, "skill": 0.534, "combined": 0.5599, "coverage": 1, "pending": 0, "n": 6}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 6

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Arts · skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Count is contributing benchmarks, not requests. Source area context: {"id": "arts", "raw": 0.7016, "skill": 0.534, "combined": 0.5599, "coverage": 1, "pending": 0, "n": 6}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 6

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Arts · combined

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Count is contributing benchmarks, not requests. Source area context: {"id": "arts", "raw": 0.7016, "skill": 0.534, "combined": 0.5599, "coverage": 1, "pending": 0, "n": 6}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 6

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · ECE

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. 32 benchmarks with a per-field right or wrong answer (ToolRet, BRIGHT and RouterBench excluded), each weighted equally; a 1-in-6 sample of requests; confidence is the probability placed on the chosen option. Rows an entrant trained on count as wrong here too. Contributing benchmarks 31.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 34411

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Brier

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. 32 benchmarks with a per-field right or wrong answer (ToolRet, BRIGHT and RouterBench excluded), each weighted equally; a 1-in-6 sample of requests; confidence is the probability placed on the chosen option. Rows an entrant trained on count as wrong here too. Contributing benchmarks 31.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 34411

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Sample accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. 32 benchmarks with a per-field right or wrong answer (ToolRet, BRIGHT and RouterBench excluded), each weighted equally; a 1-in-6 sample of requests; confidence is the probability placed on the chosen option. Rows an entrant trained on count as wrong here too. Contributing benchmarks 31.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 34411

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Mean chosen confidence

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. 32 benchmarks with a per-field right or wrong answer (ToolRet, BRIGHT and RouterBench excluded), each weighted equally; a 1-in-6 sample of requests; confidence is the probability placed on the chosen option. Rows an entrant trained on count as wrong here too. Contributing benchmarks 31.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 34411

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Wrong among answers at ≥95% confidence

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. 32 benchmarks with a per-field right or wrong answer (ToolRet, BRIGHT and RouterBench excluded), each weighted equally; a 1-in-6 sample of requests; confidence is the probability placed on the chosen option. Rows an entrant trained on count as wrong here too. Contributing benchmarks 31.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 34411

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Knowledge · ECE

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Area-specific item counts unpublished; do not use whole-sample count.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Knowledge · Sample accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Area-specific item counts unpublished; do not use whole-sample count.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Knowledge · Mean chosen confidence

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Area-specific item counts unpublished; do not use whole-sample count.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Language · ECE

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Area-specific item counts unpublished; do not use whole-sample count.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Language · Sample accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Area-specific item counts unpublished; do not use whole-sample count.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Language · Mean chosen confidence

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Area-specific item counts unpublished; do not use whole-sample count.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Retrieval · ECE

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Area-specific item counts unpublished; do not use whole-sample count.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Retrieval · Sample accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Area-specific item counts unpublished; do not use whole-sample count.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Retrieval · Mean chosen confidence

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Area-specific item counts unpublished; do not use whole-sample count.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Tools · ECE

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Area-specific item counts unpublished; do not use whole-sample count.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Tools · Sample accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Area-specific item counts unpublished; do not use whole-sample count.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Tools · Mean chosen confidence

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Area-specific item counts unpublished; do not use whole-sample count.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Arts · ECE

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Area-specific item counts unpublished; do not use whole-sample count.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Arts · Sample accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Area-specific item counts unpublished; do not use whole-sample count.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Arts · Mean chosen confidence

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Area-specific item counts unpublished; do not use whole-sample count.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Confidence bin 0–0.1 · Sample weight

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Reliability bin; source weighted fraction, not sample count. Lower boundary 0.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Confidence bin 0–0.1 · Mean confidence

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Reliability bin; source weighted fraction, not sample count. Lower boundary 0.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Confidence bin 0–0.1 · Sample accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Reliability bin; source weighted fraction, not sample count. Lower boundary 0.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Confidence bin 0.1–0.2 · Sample weight

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Reliability bin; source weighted fraction, not sample count. Lower boundary 0.1.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Confidence bin 0.1–0.2 · Mean confidence

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Reliability bin; source weighted fraction, not sample count. Lower boundary 0.1.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Confidence bin 0.1–0.2 · Sample accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Reliability bin; source weighted fraction, not sample count. Lower boundary 0.1.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Confidence bin 0.2–0.3 · Sample weight

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Reliability bin; source weighted fraction, not sample count. Lower boundary 0.2.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Confidence bin 0.2–0.3 · Mean confidence

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Reliability bin; source weighted fraction, not sample count. Lower boundary 0.2.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Confidence bin 0.2–0.3 · Sample accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Reliability bin; source weighted fraction, not sample count. Lower boundary 0.2.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Confidence bin 0.3–0.4 · Sample weight

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Reliability bin; source weighted fraction, not sample count. Lower boundary 0.3.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Confidence bin 0.3–0.4 · Mean confidence

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Reliability bin; source weighted fraction, not sample count. Lower boundary 0.3.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Confidence bin 0.3–0.4 · Sample accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Reliability bin; source weighted fraction, not sample count. Lower boundary 0.3.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Confidence bin 0.4–0.5 · Sample weight

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Reliability bin; source weighted fraction, not sample count. Lower boundary 0.4.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Confidence bin 0.4–0.5 · Mean confidence

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Reliability bin; source weighted fraction, not sample count. Lower boundary 0.4.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Confidence bin 0.4–0.5 · Sample accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Reliability bin; source weighted fraction, not sample count. Lower boundary 0.4.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Confidence bin 0.5–0.6 · Sample weight

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Reliability bin; source weighted fraction, not sample count. Lower boundary 0.5.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Confidence bin 0.5–0.6 · Mean confidence

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Reliability bin; source weighted fraction, not sample count. Lower boundary 0.5.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Confidence bin 0.5–0.6 · Sample accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Reliability bin; source weighted fraction, not sample count. Lower boundary 0.5.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Confidence bin 0.6–0.7 · Sample weight

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Reliability bin; source weighted fraction, not sample count. Lower boundary 0.6.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Confidence bin 0.6–0.7 · Mean confidence

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Reliability bin; source weighted fraction, not sample count. Lower boundary 0.6.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Confidence bin 0.6–0.7 · Sample accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Reliability bin; source weighted fraction, not sample count. Lower boundary 0.6.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Confidence bin 0.7–0.8 · Sample weight

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Reliability bin; source weighted fraction, not sample count. Lower boundary 0.7.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Confidence bin 0.7–0.8 · Mean confidence

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Reliability bin; source weighted fraction, not sample count. Lower boundary 0.7.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Confidence bin 0.7–0.8 · Sample accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Reliability bin; source weighted fraction, not sample count. Lower boundary 0.7.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Confidence bin 0.8–0.9 · Sample weight

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Reliability bin; source weighted fraction, not sample count. Lower boundary 0.8.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Confidence bin 0.8–0.9 · Mean confidence

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Reliability bin; source weighted fraction, not sample count. Lower boundary 0.8.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Confidence bin 0.8–0.9 · Sample accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Reliability bin; source weighted fraction, not sample count. Lower boundary 0.8.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Confidence bin 0.9–1.0 · Sample weight

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Reliability bin; source weighted fraction, not sample count. Lower boundary 0.9.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Confidence bin 0.9–1.0 · Mean confidence

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Reliability bin; source weighted fraction, not sample count. Lower boundary 0.9.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text calibration 0.4 · Confidence bin 0.9–1.0 · Sample accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Reliability bin; source weighted fraction, not sample count. Lower boundary 0.9.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Local median latency

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Latency is measured on its own, after scoring: every entrant runs the same 750 rows (stratified by benchmark and prompt length across the 0.2 scope, plus 10 untimed warm-up rows) as one process sending one request at a time on 1 x NVIDIA RTX PRO 6000, using the fastest inference path the model or its code supports. The figure is the wall time of one row, request in to all answers out, measured the same way for in-process and HTTP engines; board medians and p95s come from this run, not from the scoring runs. Entrant-specific serving/hardware caveats in configuration take precedence; HTTPS Jev is not isolated local inference.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Local p95 latency

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Latency is measured on its own, after scoring: every entrant runs the same 750 rows (stratified by benchmark and prompt length across the 0.2 scope, plus 10 untimed warm-up rows) as one process sending one request at a time on 1 x NVIDIA RTX PRO 6000, using the fastest inference path the model or its code supports. The figure is the wall time of one row, request in to all answers out, measured the same way for in-process and HTTP engines; board medians and p95s come from this run, not from the scoring runs. Entrant-specific serving/hardware caveats in configuration take precedence; HTTPS Jev is not isolated local inference.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Local mean latency

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Latency is measured on its own, after scoring: every entrant runs the same 750 rows (stratified by benchmark and prompt length across the 0.2 scope, plus 10 untimed warm-up rows) as one process sending one request at a time on 1 x NVIDIA RTX PRO 6000, using the fastest inference path the model or its code supports. The figure is the wall time of one row, request in to all answers out, measured the same way for in-process and HTTP engines; board medians and p95s come from this run, not from the scoring runs. Entrant-specific serving/hardware caveats in configuration take precedence; HTTPS Jev is not isolated local inference.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · BFCL · case exact accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.9793, "jev": null, "metric": "case exact accuracy", "requests": 1694, "answered": 1694, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 1, "dataset": "BFCL", "short": "BFCL", "metric": "case exact accuracy", "explainer": "Says yes or no for each supplied function: call it for this request? All must be right.", "baseline": {"value": 0.2592, "label": "Random whole-case guesses"}, "lower": false, "range": null, "cases": 1694, "group": "tools", "in_suite": true, "interactive": false, "references": [{"kind": "paper", "url": "https://arxiv.org/abs/2305.15334"}, {"kind": "code", "url": "https://github.com/gorilla-llm/gorilla"}], "gold": 1.2, "private": true}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 1694

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · BFCL · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native case exact accuracy. Source normalization: {"raw": 0.9793, "skill": 0.9721, "coverage": 1, "random": 0.2592, "tracks": [], "combined": 0.9602, "private_eq": 0.9701}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 1694

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · BFCL · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native case exact accuracy. Source normalization: {"raw": 0.9793, "skill": 0.9721, "coverage": 1, "random": 0.2592, "tracks": [], "combined": 0.9602, "private_eq": 0.9701}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 1694

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · BFCL · Combined public/private component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native case exact accuracy. Source normalization: {"raw": 0.9793, "skill": 0.9721, "coverage": 1, "random": 0.2592, "tracks": [], "combined": 0.9602, "private_eq": 0.9701}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · BFCL · Equated private same-skill component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native case exact accuracy. Source normalization: {"raw": 0.9793, "skill": 0.9721, "coverage": 1, "random": 0.2592, "tracks": [], "combined": 0.9602, "private_eq": 0.9701}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · ToolRet · nDCG@10

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.6971, "jev": null, "metric": "nDCG@10", "requests": 685, "answered": 685, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 2, "dataset": "ToolRet", "short": "ToolRet", "metric": "nDCG@10", "explainer": "Rates how useful each of 32 search-hit tools is for a request; ranked by nDCG@10, on requests where at least one of the 32 is relevant.", "baseline": {"value": 0.1341, "label": "Random ranking, answerable queries", "previous": 0.0937}, "lower": false, "range": null, "cases": 685, "group": "tools", "in_suite": true, "interactive": false, "references": [{"kind": "paper", "url": "https://arxiv.org/abs/2503.01763"}, {"kind": "dataset", "url": "https://huggingface.co/datasets/mangopy/ToolRet-Queries"}], "private": true}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 685

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · ToolRet · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native nDCG@10. Source normalization: {"raw": 0.6971, "skill": 0.6503, "coverage": 1, "random": 0.1341, "tracks": [], "combined": 0.7038, "private_eq": 0.7456}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 685

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · ToolRet · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native nDCG@10. Source normalization: {"raw": 0.6971, "skill": 0.6503, "coverage": 1, "random": 0.1341, "tracks": [], "combined": 0.7038, "private_eq": 0.7456}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 685

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · ToolRet · Combined public/private component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native nDCG@10. Source normalization: {"raw": 0.6971, "skill": 0.6503, "coverage": 1, "random": 0.1341, "tracks": [], "combined": 0.7038, "private_eq": 0.7456}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · ToolRet · Equated private same-skill component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native nDCG@10. Source normalization: {"raw": 0.6971, "skill": 0.6503, "coverage": 1, "random": 0.1341, "tracks": [], "combined": 0.7038, "private_eq": 0.7456}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · API-Bank · accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.8839, "jev": null, "metric": "accuracy", "requests": 508, "answered": 508, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 3, "dataset": "API-Bank", "short": "API-Bank", "metric": "accuracy", "explainer": "Reads an assistant conversation and picks the next API to call from a catalog of 53.", "baseline": {"value": 0.0189, "label": "Uniform random choice"}, "lower": false, "range": null, "cases": 508, "group": "tools", "in_suite": true, "interactive": false, "references": [{"kind": "paper", "url": "https://arxiv.org/abs/2304.08244"}, {"kind": "code", "url": "https://github.com/AlibabaResearch/DAMO-ConvAI/tree/main/api-bank"}], "gold": 1.2, "private": true}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 508

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · API-Bank · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.8839, "skill": 0.8817, "coverage": 1, "random": 0.0189, "tracks": [], "combined": 0.9224, "private_eq": 0.9237}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 508

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · API-Bank · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.8839, "skill": 0.8817, "coverage": 1, "random": 0.0189, "tracks": [], "combined": 0.9224, "private_eq": 0.9237}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 508

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · API-Bank · Combined public/private component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.8839, "skill": 0.8817, "coverage": 1, "random": 0.0189, "tracks": [], "combined": 0.9224, "private_eq": 0.9237}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · API-Bank · Equated private same-skill component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.8839, "skill": 0.8817, "coverage": 1, "random": 0.0189, "tracks": [], "combined": 0.9224, "private_eq": 0.9237}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · BANKING77 · macro-F1

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.9032, "jev": null, "metric": "macro-F1", "requests": 3080, "answered": 3080, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 4, "dataset": "BANKING77", "short": "BANKING77", "metric": "macro-F1", "explainer": "Picks the intent of a bank customer's message, e.g. \"How do I locate my card?\", from 77.", "baseline": {"value": 0.0127, "label": "Random choices · F1"}, "lower": false, "range": null, "cases": 3080, "group": "retrieval", "in_suite": true, "interactive": false, "references": [{"kind": "paper", "url": "https://arxiv.org/abs/2003.04807"}, {"kind": "dataset", "url": "https://github.com/PolyAI-LDN/task-specific-datasets"}], "gold": 1.2, "private": true}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 3080

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · BANKING77 · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native macro-F1. Source normalization: {"raw": 0.9032, "skill": 0.902, "coverage": 1, "random": 0.0127, "tracks": [], "combined": 0.8887, "private_eq": 0.8904}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 3080

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · BANKING77 · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native macro-F1. Source normalization: {"raw": 0.9032, "skill": 0.902, "coverage": 1, "random": 0.0127, "tracks": [], "combined": 0.8887, "private_eq": 0.8904}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 3080

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · BANKING77 · Combined public/private component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native macro-F1. Source normalization: {"raw": 0.9032, "skill": 0.902, "coverage": 1, "random": 0.0127, "tracks": [], "combined": 0.8887, "private_eq": 0.8904}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · BANKING77 · Equated private same-skill component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native macro-F1. Source normalization: {"raw": 0.9032, "skill": 0.902, "coverage": 1, "random": 0.0127, "tracks": [], "combined": 0.8887, "private_eq": 0.8904}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · CLINC150+OOS · macro-F1

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.9561, "jev": null, "metric": "macro-F1", "requests": 5500, "answered": 5500, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 5, "dataset": "CLINC150+OOS", "short": "CLINC150", "metric": "macro-F1", "explainer": "Reads a short assistant request and picks its intent from 150, or \"out of scope\" when none fits.", "baseline": {"value": 0.006, "label": "Random choices · F1"}, "lower": false, "range": null, "cases": 5500, "group": "retrieval", "in_suite": true, "interactive": false, "references": [{"kind": "dataset", "url": "https://github.com/clinc/oos-eval"}], "gold": 1.2, "private": true}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 5500

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · CLINC150+OOS · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native macro-F1. Source normalization: {"raw": 0.9561, "skill": 0.9558, "coverage": 1, "random": 0.006, "tracks": [], "combined": 0.8404, "private_eq": 0.8422}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 5500

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · CLINC150+OOS · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native macro-F1. Source normalization: {"raw": 0.9561, "skill": 0.9558, "coverage": 1, "random": 0.006, "tracks": [], "combined": 0.8404, "private_eq": 0.8422}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 5500

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · CLINC150+OOS · Combined public/private component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native macro-F1. Source normalization: {"raw": 0.9561, "skill": 0.9558, "coverage": 1, "random": 0.006, "tracks": [], "combined": 0.8404, "private_eq": 0.8422}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · CLINC150+OOS · Equated private same-skill component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native macro-F1. Source normalization: {"raw": 0.9561, "skill": 0.9558, "coverage": 1, "random": 0.006, "tracks": [], "combined": 0.8404, "private_eq": 0.8422}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · RouterBench · selected quality (quality objective)

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.7967, "jev": null, "metric": "selected quality (quality objective)", "requests": 10000, "answered": 10000, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 6, "dataset": "RouterBench", "short": "RouterBench", "metric": "selected quality (quality objective)", "explainer": "Picks which of 11 AI models should answer a prompt; scored by its real answer.", "baseline": {"value": 0.5735, "label": "Random model selection"}, "lower": false, "range": null, "cases": 5001, "group": "tools", "in_suite": true, "interactive": false, "index_note": "Not in the index: the prompt reveals that GPT-4 is the best route, so the top models all pick GPT-4 and score within a fraction of a point of each other. Scores stay visible.", "references": [{"kind": "paper", "url": "https://arxiv.org/abs/2403.12031"}, {"kind": "dataset", "url": "https://huggingface.co/datasets/withmartian/routerbench"}]}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 10000

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · RouterBench · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native selected quality (quality objective). Source normalization: {"raw": 0.7967, "skill": 0.5233, "coverage": 1, "random": 0.5735, "tracks": []}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 10000

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · RouterBench · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native selected quality (quality objective). Source normalization: {"raw": 0.7967, "skill": 0.5233, "coverage": 1, "random": 0.5735, "tracks": []}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 10000

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Home appliance simulator · case exact accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.7273, "jev": null, "metric": "case exact accuracy", "requests": 88, "answered": 88, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 9, "dataset": "Home appliance simulator", "short": "Home appliances", "metric": "case exact accuracy", "explainer": "Given a simulated home and a command, answers ~18 questions on handling it; all must be right. Duplicate and dev-set rows removed.", "baseline": {"value": 0, "label": "Random whole-case guesses"}, "lower": false, "range": null, "cases": 88, "group": "tools", "in_suite": true, "interactive": false, "references": [{"kind": "code", "url": "https://github.com/apolinario/decision-index/blob/main/decision_index/suite/build/home_appliance.py"}], "private": true}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 88

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Home appliance simulator · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native case exact accuracy. Source normalization: {"raw": 0.7273, "skill": 0.7273, "coverage": 1, "random": 0, "tracks": [], "combined": 0.7051, "private_eq": 0.7075}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 88

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Home appliance simulator · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native case exact accuracy. Source normalization: {"raw": 0.7273, "skill": 0.7273, "coverage": 1, "random": 0, "tracks": [], "combined": 0.7051, "private_eq": 0.7075}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 88

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Home appliance simulator · Combined public/private component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native case exact accuracy. Source normalization: {"raw": 0.7273, "skill": 0.7273, "coverage": 1, "random": 0, "tracks": [], "combined": 0.7051, "private_eq": 0.7075}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Home appliance simulator · Equated private same-skill component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native case exact accuracy. Source normalization: {"raw": 0.7273, "skill": 0.7273, "coverage": 1, "random": 0, "tracks": [], "combined": 0.7051, "private_eq": 0.7075}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · SGD/SGD-X · macro-F1

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.5764, "jev": null, "metric": "macro-F1", "requests": 2500, "answered": 2500, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 10, "dataset": "SGD/SGD-X", "short": "SGD", "metric": "macro-F1", "explainer": "Reads a service dialogue so far and picks the intent the user is pursuing now, or NONE.", "baseline": {"value": 0.399, "label": "Random choices · F1"}, "lower": false, "range": null, "cases": 2500, "group": "retrieval", "in_suite": true, "interactive": false, "index_note": "Not in the index: build bug, pending re-run. The builder gave every SGD row the whole dialogue, including the turns after the one being classified. Scores stay visible; SGD returns to the index once re-run with the fixed builder.", "references": [{"kind": "paper", "url": "https://arxiv.org/abs/1909.05855"}, {"kind": "dataset", "url": "https://github.com/google-research-datasets/dstc8-schema-guided-dialogue"}]}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 2500

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · SGD/SGD-X · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native macro-F1. Source normalization: {"raw": 0.5764, "skill": 0.2952, "coverage": 1, "random": 0.399, "tracks": []}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 2500

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · SGD/SGD-X · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native macro-F1. Source normalization: {"raw": 0.5764, "skill": 0.2952, "coverage": 1, "random": 0.399, "tracks": []}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 2500

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · ContractNLI · macro-F1

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.8392, "jev": null, "metric": "macro-F1", "requests": 123, "answered": 123, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 11, "dataset": "ContractNLI", "short": "ContractNLI", "metric": "macro-F1", "explainer": "Says whether a full NDA supports, contradicts or omits each of 17 fixed statements.", "baseline": {"value": 0.3085, "label": "Random choices · F1"}, "lower": false, "range": null, "cases": 123, "group": "language", "in_suite": true, "interactive": false, "references": [{"kind": "paper", "url": "https://aclanthology.org/2021.findings-emnlp.164"}, {"kind": "dataset", "url": "https://stanfordnlp.github.io/contract-nli/"}]}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 123

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · ContractNLI · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native macro-F1. Source normalization: {"raw": 0.8392, "skill": 0.7674, "coverage": 1, "random": 0.3085, "tracks": []}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 123

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · ContractNLI · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native macro-F1. Source normalization: {"raw": 0.8392, "skill": 0.7674, "coverage": 1, "random": 0.3085, "tracks": []}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 123

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · ANLI · macro-F1

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.7709, "jev": null, "metric": "macro-F1", "requests": 3200, "answered": 3200, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 12, "dataset": "ANLI", "short": "ANLI", "metric": "macro-F1", "explainer": "Decides whether a claim follows from, contradicts, or is neutral to a premise written to fool AI models.", "baseline": {"value": 0.3324, "label": "Random choices · F1"}, "lower": false, "range": null, "cases": 3200, "group": "language", "in_suite": true, "interactive": false, "references": [{"kind": "paper", "url": "https://arxiv.org/abs/1910.14599"}, {"kind": "dataset", "url": "https://huggingface.co/datasets/facebook/anli"}], "gold": 1.2, "private": true}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 3200

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · ANLI · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native macro-F1. Source normalization: {"raw": 0.7709, "skill": 0.6568, "coverage": 1, "random": 0.3324, "tracks": [], "combined": 0.5089, "private_eq": 0.6752}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 3200

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · ANLI · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native macro-F1. Source normalization: {"raw": 0.7709, "skill": 0.6568, "coverage": 1, "random": 0.3324, "tracks": [], "combined": 0.5089, "private_eq": 0.6752}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 3200

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · ANLI · Combined public/private component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native macro-F1. Source normalization: {"raw": 0.7709, "skill": 0.6568, "coverage": 1, "random": 0.3324, "tracks": [], "combined": 0.5089, "private_eq": 0.6752}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · ANLI · Equated private same-skill component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native macro-F1. Source normalization: {"raw": 0.7709, "skill": 0.6568, "coverage": 1, "random": 0.3324, "tracks": [], "combined": 0.5089, "private_eq": 0.6752}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · BPoMP · accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.9532, "jev": null, "metric": "accuracy", "requests": 5000, "answered": 5000, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 20, "dataset": "BPoMP", "short": "BPoMP", "metric": "accuracy", "explainer": "Given two versions of a limerick, one with its rhyme, meter or sense broken, picks the original.", "baseline": {"value": 0.5, "label": "Uniform random choice"}, "lower": false, "range": null, "cases": 811, "group": "arts", "in_suite": true, "interactive": false, "references": [{"kind": "dataset", "url": "https://zenodo.org/records/7299879"}], "private": true}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 5000

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · BPoMP · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.9546, "skill": 0.9092, "coverage": 1, "random": 0.5, "tracks": [], "combined": 0.9577, "private_eq": 0.9786}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 5000

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · BPoMP · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.9546, "skill": 0.9092, "coverage": 1, "random": 0.5, "tracks": [], "combined": 0.9577, "private_eq": 0.9786}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 5000

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · BPoMP · Combined public/private component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.9546, "skill": 0.9092, "coverage": 1, "random": 0.5, "tracks": [], "combined": 0.9577, "private_eq": 0.9786}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · BPoMP · Equated private same-skill component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.9546, "skill": 0.9092, "coverage": 1, "random": 0.5, "tracks": [], "combined": 0.9577, "private_eq": 0.9786}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Humicroedit · accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.6454, "jev": null, "metric": "accuracy", "requests": 2628, "answered": 2628, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 21, "dataset": "Humicroedit", "short": "Humicroedit", "metric": "accuracy", "explainer": "Given two one-word edits of a news headline, picks the one people rated funnier.", "baseline": {"value": 0.5, "label": "Uniform random choice"}, "lower": false, "range": null, "cases": 2628, "group": "arts", "in_suite": true, "interactive": false, "references": [{"kind": "paper", "url": "https://arxiv.org/abs/1906.00274"}, {"kind": "dataset", "url": "https://cs.rochester.edu/u/nhossain/humicroedit.html"}], "private": true}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 2628

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Humicroedit · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.6454, "skill": 0.2907, "coverage": 1, "random": 0.5, "tracks": [], "combined": 0.2264, "private_eq": 0.6171}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 2628

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Humicroedit · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.6454, "skill": 0.2907, "coverage": 1, "random": 0.5, "tracks": [], "combined": 0.2264, "private_eq": 0.6171}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 2628

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Humicroedit · Combined public/private component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.6454, "skill": 0.2907, "coverage": 1, "random": 0.5, "tracks": [], "combined": 0.2264, "private_eq": 0.6171}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Humicroedit · Equated private same-skill component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.6454, "skill": 0.2907, "coverage": 1, "random": 0.5, "tracks": [], "combined": 0.2264, "private_eq": 0.6171}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · POP909-CL · accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.6305, "jev": null, "metric": "accuracy", "requests": 2000, "answered": 2000, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 22, "dataset": "POP909-CL", "short": "POP909", "metric": "accuracy", "explainer": "Given the notes around a beat in a pop song (not audio), picks the chord playing from 129.", "baseline": {"value": 0.0078, "label": "Uniform random choice"}, "lower": false, "range": null, "cases": 2000, "group": "arts", "in_suite": true, "interactive": false, "references": [{"kind": "paper", "url": "https://arxiv.org/abs/2510.06528"}, {"kind": "dataset", "url": "https://github.com/AndyWeasley2004/POP909-CL-Dataset"}], "private": true}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 2000

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · POP909-CL · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.6282, "skill": 0.6253, "coverage": 1, "random": 0.0078, "tracks": [], "combined": 0.545, "private_eq": 0.5527}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 2000

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · POP909-CL · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.6282, "skill": 0.6253, "coverage": 1, "random": 0.0078, "tracks": [], "combined": 0.545, "private_eq": 0.5527}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 2000

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · POP909-CL · Combined public/private component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.6282, "skill": 0.6253, "coverage": 1, "random": 0.0078, "tracks": [], "combined": 0.545, "private_eq": 0.5527}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · POP909-CL · Equated private same-skill component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.6282, "skill": 0.6253, "coverage": 1, "random": 0.0078, "tracks": [], "combined": 0.545, "private_eq": 0.5527}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · cfcolor · accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.6624, "jev": null, "metric": "accuracy", "requests": 5000, "answered": 5000, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 23, "dataset": "cfcolor", "short": "cfcolor", "metric": "accuracy", "explainer": "Given a person's earlier ratings of color palettes, predicts which of two new palettes they rated higher.", "baseline": {"value": 0.5, "label": "Uniform random choice"}, "lower": false, "range": null, "cases": 5000, "group": "arts", "in_suite": true, "interactive": false, "references": [{"kind": "dataset", "url": "https://www.dgp.toronto.edu/~donovan/cfcolor/"}]}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 5000

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · cfcolor · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.6681, "skill": 0.3363, "coverage": 1, "random": 0.5, "tracks": []}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 5000

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · cfcolor · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.6681, "skill": 0.3363, "coverage": 1, "random": 0.5, "tracks": []}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 5000

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · MMLU · accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.8786, "jev": null, "metric": "accuracy", "requests": 14033, "answered": 14033, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 24, "dataset": "MMLU", "short": "MMLU", "metric": "accuracy", "explainer": "Four-option questions across 57 school and professional subjects, from law to algebra.", "baseline": {"value": 0.25, "label": "Uniform random choice"}, "lower": false, "range": null, "cases": 14042, "group": "knowledge", "in_suite": true, "interactive": false, "references": [{"kind": "dataset", "url": "https://huggingface.co/datasets/cais/mmlu"}]}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 14033

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · MMLU · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.8786, "skill": 0.8381, "coverage": 1, "random": 0.25, "tracks": []}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 14033

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · MMLU · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.8786, "skill": 0.8381, "coverage": 1, "random": 0.25, "tracks": []}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 14033

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · GPQA Diamond · accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.5612, "jev": null, "metric": "accuracy", "requests": 196, "answered": 196, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 25, "dataset": "GPQA Diamond", "short": "GPQA Diamond", "metric": "accuracy", "explainer": "198 hard four-option science questions written by PhD experts in physics, chemistry and biology.", "baseline": {"value": 0.25, "label": "Uniform random choice"}, "lower": false, "range": null, "cases": 198, "group": "knowledge", "in_suite": true, "interactive": false, "references": [{"kind": "dataset", "url": "https://github.com/idavidrein/gpqa"}], "gold": 1.2, "private": true}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 196

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · GPQA Diamond · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.5612, "skill": 0.415, "coverage": 1, "random": 0.25, "tracks": [], "private_eq": 0.5179, "combined": 0.3508}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 196

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · GPQA Diamond · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.5612, "skill": 0.415, "coverage": 1, "random": 0.25, "tracks": [], "private_eq": 0.5179, "combined": 0.3508}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 196

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · GPQA Diamond · Combined public/private component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.5612, "skill": 0.415, "coverage": 1, "random": 0.25, "tracks": [], "private_eq": 0.5179, "combined": 0.3508}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · GPQA Diamond · Equated private same-skill component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.5612, "skill": 0.415, "coverage": 1, "random": 0.25, "tracks": [], "private_eq": 0.5179, "combined": 0.3508}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · ARC-Easy · accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.9924, "jev": null, "metric": "accuracy", "requests": 2376, "answered": 2376, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 26, "dataset": "ARC-Easy", "short": "ARC-Easy", "metric": "accuracy", "explainer": "Grade-school science multiple choice, the easier set.", "baseline": {"value": 0.2502, "label": "Uniform random choice"}, "lower": false, "range": null, "cases": 2376, "group": "knowledge", "in_suite": true, "interactive": false, "references": [{"kind": "dataset", "url": "https://huggingface.co/datasets/allenai/ai2_arc"}]}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 2376

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · ARC-Easy · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.9924, "skill": 0.9899, "coverage": 1, "random": 0.2502, "tracks": []}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 2376

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · ARC-Easy · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.9924, "skill": 0.9899, "coverage": 1, "random": 0.2502, "tracks": []}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 2376

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · ARC-Challenge · accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.9727, "jev": null, "metric": "accuracy", "requests": 1172, "answered": 1172, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 27, "dataset": "ARC-Challenge", "short": "ARC-Challenge", "metric": "accuracy", "explainer": "Grade-school science multiple choice, the harder set that simple retrieval gets wrong.", "baseline": {"value": 0.2502, "label": "Uniform random choice"}, "lower": false, "range": null, "cases": 1172, "group": "knowledge", "in_suite": true, "interactive": false, "references": [{"kind": "dataset", "url": "https://huggingface.co/datasets/allenai/ai2_arc"}]}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 1172

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · ARC-Challenge · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.9727, "skill": 0.9636, "coverage": 1, "random": 0.2502, "tracks": []}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 1172

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · ARC-Challenge · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.9727, "skill": 0.9636, "coverage": 1, "random": 0.2502, "tracks": []}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 1172

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · WinoGrande · accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.9519, "jev": null, "metric": "accuracy", "requests": 1267, "answered": 1267, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 28, "dataset": "WinoGrande", "short": "WinoGrande", "metric": "accuracy", "explainer": "Fills a blank with one of two options using common sense, e.g. who \"got the easier cases\".", "baseline": {"value": 0.5, "label": "Uniform random choice"}, "lower": false, "range": null, "cases": 1267, "group": "knowledge", "in_suite": true, "interactive": false, "references": [{"kind": "dataset", "url": "https://huggingface.co/datasets/allenai/winogrande"}], "gold": 1.2, "private": true}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 1267

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · WinoGrande · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.9519, "skill": 0.9038, "coverage": 1, "random": 0.5, "tracks": [], "combined": 0.6429, "private_eq": 0.823}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 1267

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · WinoGrande · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.9519, "skill": 0.9038, "coverage": 1, "random": 0.5, "tracks": [], "combined": 0.6429, "private_eq": 0.823}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 1267

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · WinoGrande · Combined public/private component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.9519, "skill": 0.9038, "coverage": 1, "random": 0.5, "tracks": [], "combined": 0.6429, "private_eq": 0.823}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · WinoGrande · Equated private same-skill component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.9519, "skill": 0.9038, "coverage": 1, "random": 0.5, "tracks": [], "combined": 0.6429, "private_eq": 0.823}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · HellaSwag · accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.9794, "jev": null, "metric": "accuracy", "requests": 10042, "answered": 10042, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 29, "dataset": "HellaSwag", "short": "HellaSwag", "metric": "accuracy", "explainer": "Picks the most sensible of four continuations of an everyday scene.", "baseline": {"value": 0.25, "label": "Uniform random choice"}, "lower": false, "range": null, "cases": 10042, "group": "language", "in_suite": true, "interactive": false, "references": [{"kind": "dataset", "url": "https://huggingface.co/datasets/Rowan/hellaswag"}], "gold": 1.2, "private": true}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 10042

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · HellaSwag · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.9794, "skill": 0.9725, "coverage": 1, "random": 0.25, "tracks": [], "combined": 1.001, "private_eq": 1}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 10042

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · HellaSwag · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.9794, "skill": 0.9725, "coverage": 1, "random": 0.25, "tracks": [], "combined": 1.001, "private_eq": 1}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 10042

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · HellaSwag · Combined public/private component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.9794, "skill": 0.9725, "coverage": 1, "random": 0.25, "tracks": [], "combined": 1.001, "private_eq": 1}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · HellaSwag · Equated private same-skill component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.9794, "skill": 0.9725, "coverage": 1, "random": 0.25, "tracks": [], "combined": 1.001, "private_eq": 1}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · GSM8K · accuracy (4- and 10-choice)

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.8336, "jev": 0.7301, "metric": "accuracy (4- and 10-choice)", "requests": 2638, "answered": 2638, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 30, "dataset": "GSM8K", "short": "GSM8K", "metric": "accuracy", "explainer": "Picks a math word problem's answer from 4, and separately 10, numbers, with no working. Rebuilt for v0.3.", "baseline": {"value": 0.175, "label": "Uniform random choice"}, "lower": false, "range": null, "cases": 2638, "group": "knowledge", "in_suite": true, "interactive": false, "references": [{"kind": "dataset", "url": "https://huggingface.co/datasets/openai/gsm8k"}], "private": true}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 2638

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · GSM8K · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.8336, "skill": 0.8003, "coverage": 1, "random": 0.175, "tracks": [{"track": "4-choice", "raw": 0.8666, "random": 0.25}, {"track": "10-choice", "raw": 0.8006, "random": 0.1}], "combined": 0.7746, "private_eq": 0.8153}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 2638

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · GSM8K · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.8336, "skill": 0.8003, "coverage": 1, "random": 0.175, "tracks": [{"track": "4-choice", "raw": 0.8666, "random": 0.25}, {"track": "10-choice", "raw": 0.8006, "random": 0.1}], "combined": 0.7746, "private_eq": 0.8153}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 2638

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · GSM8K · Combined public/private component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.8336, "skill": 0.8003, "coverage": 1, "random": 0.175, "tracks": [{"track": "4-choice", "raw": 0.8666, "random": 0.25}, {"track": "10-choice", "raw": 0.8006, "random": 0.1}], "combined": 0.7746, "private_eq": 0.8153}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · GSM8K · Equated private same-skill component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.8336, "skill": 0.8003, "coverage": 1, "random": 0.175, "tracks": [{"track": "4-choice", "raw": 0.8666, "random": 0.25}, {"track": "10-choice", "raw": 0.8006, "random": 0.1}], "combined": 0.7746, "private_eq": 0.8153}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · GSM8K · 4-choice

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Published separate track; sample count unspecified. {"track": "4-choice", "raw": 0.8666, "random": 0.25}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · GSM8K · 10-choice

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Published separate track; sample count unspecified. {"track": "10-choice", "raw": 0.8006, "random": 0.1}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · ChessBench · accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.2454, "jev": null, "metric": "accuracy", "requests": 5000, "answered": 5000, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 31, "dataset": "ChessBench", "short": "ChessBench", "metric": "accuracy", "explainer": "Given a chess position and every legal move, picks the strongest one; tied best moves count.", "baseline": {"value": 0.0819, "label": "Uniform random choice"}, "lower": false, "range": null, "cases": 5000, "group": "knowledge", "in_suite": true, "interactive": false, "references": [{"kind": "paper", "url": "https://arxiv.org/abs/2402.04494"}, {"kind": "code", "url": "https://github.com/google-deepmind/searchless_chess"}], "private": true}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 5000

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · ChessBench · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.2454, "skill": 0.1781, "coverage": 1, "random": 0.0819, "tracks": [], "combined": 0.0723, "private_eq": 0.1571}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 5000

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · ChessBench · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.2454, "skill": 0.1781, "coverage": 1, "random": 0.0819, "tracks": [], "combined": 0.0723, "private_eq": 0.1571}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 5000

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · ChessBench · Combined public/private component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.2454, "skill": 0.1781, "coverage": 1, "random": 0.0819, "tracks": [], "combined": 0.0723, "private_eq": 0.1571}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · ChessBench · Equated private same-skill component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.2454, "skill": 0.1781, "coverage": 1, "random": 0.0819, "tracks": [], "combined": 0.0723, "private_eq": 0.1571}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · MuSR · accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.6968, "jev": null, "metric": "accuracy", "requests": 752, "answered": 752, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 32, "dataset": "MuSR", "short": "MuSR", "metric": "accuracy", "explainer": "Reads a long story and answers one question needing clues from across it, e.g. who the murderer is.", "baseline": {"value": 0.371, "label": "Uniform random choice"}, "lower": false, "range": null, "cases": 752, "group": "knowledge", "in_suite": true, "interactive": false, "references": [{"kind": "paper", "url": "https://arxiv.org/abs/2310.16049"}, {"kind": "dataset", "url": "https://github.com/Zayne-sprague/MuSR"}], "private": true}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 752

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · MuSR · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.6968, "skill": 0.518, "coverage": 1, "random": 0.371, "tracks": [], "combined": 0.3589, "private_eq": 0.6007}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 752

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · MuSR · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.6968, "skill": 0.518, "coverage": 1, "random": 0.371, "tracks": [], "combined": 0.3589, "private_eq": 0.6007}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 752

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · MuSR · Combined public/private component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.6968, "skill": 0.518, "coverage": 1, "random": 0.371, "tracks": [], "combined": 0.3589, "private_eq": 0.6007}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · MuSR · Equated private same-skill component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.6968, "skill": 0.518, "coverage": 1, "random": 0.371, "tracks": [], "combined": 0.3589, "private_eq": 0.6007}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · SATA-Bench · case exact accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.3127, "jev": null, "metric": "case exact accuracy", "requests": 1650, "answered": 1650, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 33, "dataset": "SATA-Bench", "short": "SATA-Bench", "metric": "case exact accuracy", "explainer": "Given a passage, a question and candidate answers, says yes or no to each; all must be right.", "baseline": {"value": 0.0131, "label": "Random whole-case guesses"}, "lower": false, "range": null, "cases": 1650, "group": "knowledge", "in_suite": true, "interactive": false, "references": [{"kind": "dataset", "url": "https://github.com/sata-bench/sata-bench"}], "private": true}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 1650

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · SATA-Bench · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native case exact accuracy. Source normalization: {"raw": 0.3127, "skill": 0.3036, "coverage": 1, "random": 0.0131, "tracks": [], "combined": 0.405, "private_eq": 0.4185}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 1650

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · SATA-Bench · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native case exact accuracy. Source normalization: {"raw": 0.3127, "skill": 0.3036, "coverage": 1, "random": 0.0131, "tracks": [], "combined": 0.405, "private_eq": 0.4185}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 1650

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · SATA-Bench · Combined public/private component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native case exact accuracy. Source normalization: {"raw": 0.3127, "skill": 0.3036, "coverage": 1, "random": 0.0131, "tracks": [], "combined": 0.405, "private_eq": 0.4185}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · SATA-Bench · Equated private same-skill component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native case exact accuracy. Source normalization: {"raw": 0.3127, "skill": 0.3036, "coverage": 1, "random": 0.0131, "tracks": [], "combined": 0.405, "private_eq": 0.4185}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · BRIGHT · nDCG@10

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.4988, "jev": null, "metric": "nDCG@10", "requests": 220, "answered": 220, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 36, "dataset": "BRIGHT", "short": "BRIGHT", "metric": "nDCG@10", "explainer": "Rates how helpful each of 32 search-hit documents is for a hard question; ranked by nDCG@10, on questions where at least one of the 32 is relevant.", "baseline": {"value": 0.116, "label": "Random ranking, answerable queries", "previous": 0.0448}, "lower": false, "range": null, "cases": 220, "group": "retrieval", "in_suite": true, "interactive": false, "references": [{"kind": "paper", "url": "https://arxiv.org/abs/2407.12883"}, {"kind": "dataset", "url": "https://huggingface.co/datasets/xlangai/BRIGHT"}], "gold": 1.2, "private": true}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 220

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · BRIGHT · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native nDCG@10. Source normalization: {"raw": 0.4988, "skill": 0.433, "coverage": 1, "random": 0.116, "tracks": [], "combined": 0.4147, "private_eq": 0.4876}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 220

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · BRIGHT · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native nDCG@10. Source normalization: {"raw": 0.4988, "skill": 0.433, "coverage": 1, "random": 0.116, "tracks": [], "combined": 0.4147, "private_eq": 0.4876}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 220

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · BRIGHT · Combined public/private component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native nDCG@10. Source normalization: {"raw": 0.4988, "skill": 0.433, "coverage": 1, "random": 0.116, "tracks": [], "combined": 0.4147, "private_eq": 0.4876}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · BRIGHT · Equated private same-skill component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native nDCG@10. Source normalization: {"raw": 0.4988, "skill": 0.433, "coverage": 1, "random": 0.116, "tracks": [], "combined": 0.4147, "private_eq": 0.4876}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Amazon ESCI · macro-F1

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.6273, "jev": null, "metric": "macro-F1", "requests": 5000, "answered": 5000, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 37, "dataset": "Amazon ESCI", "short": "Amazon ESCI", "metric": "macro-F1", "explainer": "Labels a product as an exact match, substitute, complement or irrelevant to a shopping query.", "baseline": {"value": 0.2027, "label": "Random choices · F1"}, "lower": false, "range": null, "cases": 5000, "group": "retrieval", "in_suite": true, "interactive": false, "references": [{"kind": "dataset", "url": "https://github.com/amazon-science/esci-data"}], "private": true}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 5000

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Amazon ESCI · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native macro-F1. Source normalization: {"raw": 0.6273, "skill": 0.5325, "coverage": 1, "random": 0.2027, "tracks": [], "combined": 0.5912, "private_eq": 0.677}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 5000

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Amazon ESCI · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native macro-F1. Source normalization: {"raw": 0.6273, "skill": 0.5325, "coverage": 1, "random": 0.2027, "tracks": [], "combined": 0.5912, "private_eq": 0.677}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 5000

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Amazon ESCI · Combined public/private component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native macro-F1. Source normalization: {"raw": 0.6273, "skill": 0.5325, "coverage": 1, "random": 0.2027, "tracks": [], "combined": 0.5912, "private_eq": 0.677}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Amazon ESCI · Equated private same-skill component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native macro-F1. Source normalization: {"raw": 0.6273, "skill": 0.5325, "coverage": 1, "random": 0.2027, "tracks": [], "combined": 0.5912, "private_eq": 0.677}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · ACOS · per-review F1

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.4115, "jev": null, "metric": "per-review F1", "requests": 1565, "answered": 1565, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 38, "dataset": "ACOS", "short": "ACOS", "metric": "per-review F1", "explainer": "Reads a restaurant or laptop review and says yes/no for each aspect-and-sentiment pair; scored by F1 per review, averaged over reviews.", "baseline": {"value": 0.031, "label": "Always yes · per-review F1", "previous": 0}, "lower": false, "range": null, "cases": 400, "group": "language", "in_suite": true, "interactive": false, "references": [{"kind": "paper", "url": "https://aclanthology.org/2021.acl-long.29"}, {"kind": "dataset", "url": "https://github.com/NUSTM/ACOS"}], "private": true}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 1565

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · ACOS · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native per-review F1. Source normalization: {"raw": 0.4115, "skill": 0.3927, "coverage": 1, "random": 0.031, "tracks": [], "combined": 0.5778, "private_eq": 0.5946}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 1565

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · ACOS · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native per-review F1. Source normalization: {"raw": 0.4115, "skill": 0.3927, "coverage": 1, "random": 0.031, "tracks": [], "combined": 0.5778, "private_eq": 0.5946}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 1565

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · ACOS · Combined public/private component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native per-review F1. Source normalization: {"raw": 0.4115, "skill": 0.3927, "coverage": 1, "random": 0.031, "tracks": [], "combined": 0.5778, "private_eq": 0.5946}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · ACOS · Equated private same-skill component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native per-review F1. Source normalization: {"raw": 0.4115, "skill": 0.3927, "coverage": 1, "random": 0.031, "tracks": [], "combined": 0.5778, "private_eq": 0.5946}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · FinEntity · macro-F1

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.9458, "jev": null, "metric": "macro-F1", "requests": 979, "answered": 979, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 39, "dataset": "FinEntity", "short": "FinEntity", "metric": "macro-F1", "explainer": "Says whether a financial news snippet is positive, negative or neutral about a named firm.", "baseline": {"value": 0.3201, "label": "Random choices · F1"}, "lower": false, "range": null, "cases": 979, "group": "language", "in_suite": true, "interactive": false, "references": [{"kind": "paper", "url": "https://aclanthology.org/2023.emnlp-main.956"}, {"kind": "dataset", "url": "https://github.com/yixuantt/FinEntity"}], "private": true}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 979

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · FinEntity · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native macro-F1. Source normalization: {"raw": 0.9458, "skill": 0.9203, "coverage": 1, "random": 0.3201, "tracks": [], "combined": 0.4394, "private_eq": 0.6225}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 979

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · FinEntity · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native macro-F1. Source normalization: {"raw": 0.9458, "skill": 0.9203, "coverage": 1, "random": 0.3201, "tracks": [], "combined": 0.4394, "private_eq": 0.6225}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 979

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · FinEntity · Combined public/private component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native macro-F1. Source normalization: {"raw": 0.9458, "skill": 0.9203, "coverage": 1, "random": 0.3201, "tracks": [], "combined": 0.4394, "private_eq": 0.6225}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · FinEntity · Equated private same-skill component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native macro-F1. Source normalization: {"raw": 0.9458, "skill": 0.9203, "coverage": 1, "random": 0.3201, "tracks": [], "combined": 0.4394, "private_eq": 0.6225}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · iSarcasmEval · Sarcasm F1 · track A, English

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.7333, "jev": null, "metric": "Sarcasm F1 \u00b7 track A, English", "requests": 4600, "answered": 4600, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 40, "dataset": "iSarcasmEval", "short": "iSarcasmEval", "metric": "Sarcasm F1 · track A, English", "explainer": "Says yes or no: is this tweet sarcastic? Scored with F1 on sarcastic English tweets (track A); other tracks listed under each result.", "baseline": {"value": 0.2227, "label": "Random choices · F1"}, "lower": false, "range": null, "cases": 4600, "group": "language", "in_suite": true, "interactive": false, "references": [{"kind": "paper", "url": "https://aclanthology.org/2022.semeval-1.111"}, {"kind": "dataset", "url": "https://github.com/iabufarha/iSarcasmEval"}], "private": true}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 4600

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · iSarcasmEval · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native Sarcasm F1 · track A, English. Source normalization: {"raw": 0.7333, "skill": 0.6569, "coverage": 1, "random": 0.2227, "tracks": [{"track": "A · Arabic", "score": 0.5603, "headline": false}, {"track": "A · English", "score": 0.7333, "headline": true}, {"track": "C · Arabic pairs", "score": 0.95, "headline": false}, {"track": "C · English pairs", "score": 0.97, "headline": false}], "combined": 0.6325, "private_eq": 0.7168}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 4600

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · iSarcasmEval · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native Sarcasm F1 · track A, English. Source normalization: {"raw": 0.7333, "skill": 0.6569, "coverage": 1, "random": 0.2227, "tracks": [{"track": "A · Arabic", "score": 0.5603, "headline": false}, {"track": "A · English", "score": 0.7333, "headline": true}, {"track": "C · Arabic pairs", "score": 0.95, "headline": false}, {"track": "C · English pairs", "score": 0.97, "headline": false}], "combined": 0.6325, "private_eq": 0.7168}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 4600

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · iSarcasmEval · Combined public/private component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native Sarcasm F1 · track A, English. Source normalization: {"raw": 0.7333, "skill": 0.6569, "coverage": 1, "random": 0.2227, "tracks": [{"track": "A · Arabic", "score": 0.5603, "headline": false}, {"track": "A · English", "score": 0.7333, "headline": true}, {"track": "C · Arabic pairs", "score": 0.95, "headline": false}, {"track": "C · English pairs", "score": 0.97, "headline": false}], "combined": 0.6325, "private_eq": 0.7168}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · iSarcasmEval · Equated private same-skill component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native Sarcasm F1 · track A, English. Source normalization: {"raw": 0.7333, "skill": 0.6569, "coverage": 1, "random": 0.2227, "tracks": [{"track": "A · Arabic", "score": 0.5603, "headline": false}, {"track": "A · English", "score": 0.7333, "headline": true}, {"track": "C · Arabic pairs", "score": 0.95, "headline": false}, {"track": "C · English pairs", "score": 0.97, "headline": false}], "combined": 0.6325, "private_eq": 0.7168}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · iSarcasmEval · A · Arabic

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Published separate track; sample count unspecified. {"track": "A · Arabic", "score": 0.5603, "headline": false}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · iSarcasmEval · C · Arabic pairs

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Published separate track; sample count unspecified. {"track": "C · Arabic pairs", "score": 0.95, "headline": false}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · iSarcasmEval · C · English pairs

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Published separate track; sample count unspecified. {"track": "C · English pairs", "score": 0.97, "headline": false}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · VAST · macro-F1

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.7834, "jev": null, "metric": "macro-F1", "requests": 3006, "answered": 3006, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 41, "dataset": "VAST", "short": "VAST", "metric": "macro-F1", "explainer": "Says whether a comment is for, against, or neutral on a given topic.", "baseline": {"value": 0.3333, "label": "Random choices · F1"}, "lower": false, "range": null, "cases": 3006, "group": "language", "in_suite": true, "interactive": false, "references": [{"kind": "dataset", "url": "https://github.com/emilyallaway/zero-shot-stance"}], "private": true}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 3006

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · VAST · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native macro-F1. Source normalization: {"raw": 0.7834, "skill": 0.6752, "coverage": 1, "random": 0.3333, "tracks": [], "combined": 0.5275, "private_eq": 0.6879}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 3006

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · VAST · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native macro-F1. Source normalization: {"raw": 0.7834, "skill": 0.6752, "coverage": 1, "random": 0.3333, "tracks": [], "combined": 0.5275, "private_eq": 0.6879}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 3006

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · VAST · Combined public/private component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native macro-F1. Source normalization: {"raw": 0.7834, "skill": 0.6752, "coverage": 1, "random": 0.3333, "tracks": [], "combined": 0.5275, "private_eq": 0.6879}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · VAST · Equated private same-skill component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native macro-F1. Source normalization: {"raw": 0.7834, "skill": 0.6752, "coverage": 1, "random": 0.3333, "tracks": [], "combined": 0.5275, "private_eq": 0.6879}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · NLI4CT · macro-F1

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.864, "jev": null, "metric": "macro-F1", "requests": 5500, "answered": 5500, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 42, "dataset": "NLI4CT", "short": "NLI4CT", "metric": "macro-F1", "explainer": "Says whether clinical-trial report sections support or contradict a statement, some deliberately altered.", "baseline": {"value": 0.486, "label": "Random choices · F1"}, "lower": false, "range": null, "cases": 5500, "group": "language", "in_suite": true, "interactive": false, "references": [{"kind": "dataset", "url": "https://github.com/ai-systems/Task-2-SemEval-2024"}], "private": true}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 5500

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · NLI4CT · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native macro-F1. Source normalization: {"raw": 0.864, "skill": 0.7354, "coverage": 1, "random": 0.486, "tracks": [], "combined": 0.9339, "private_eq": 0.9659}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 5500

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · NLI4CT · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native macro-F1. Source normalization: {"raw": 0.864, "skill": 0.7354, "coverage": 1, "random": 0.486, "tracks": [], "combined": 0.9339, "private_eq": 0.9659}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 5500

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · NLI4CT · Combined public/private component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native macro-F1. Source normalization: {"raw": 0.864, "skill": 0.7354, "coverage": 1, "random": 0.486, "tracks": [], "combined": 0.9339, "private_eq": 0.9659}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · NLI4CT · Equated private same-skill component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native macro-F1. Source normalization: {"raw": 0.864, "skill": 0.7354, "coverage": 1, "random": 0.486, "tracks": [], "combined": 0.9339, "private_eq": 0.9659}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · CRUXEval · accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.8684, "jev": null, "metric": "accuracy", "requests": 570, "answered": 570, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 43, "dataset": "CRUXEval", "short": "CRUXEval", "metric": "accuracy", "explainer": "Reads a short Python function and an input and picks its output from 2 to 10 candidates, without running it.", "baseline": {"value": 0.3697, "label": "Uniform random choice"}, "lower": false, "range": null, "cases": 570, "group": "knowledge", "in_suite": true, "interactive": false, "references": [{"kind": "paper", "url": "https://arxiv.org/abs/2401.03065"}, {"kind": "dataset", "url": "https://github.com/facebookresearch/cruxeval"}], "private": true}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 570

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · CRUXEval · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.8684, "skill": 0.7912, "coverage": 1, "random": 0.3697, "tracks": [], "combined": 0.6205, "private_eq": 0.7629}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 570

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · CRUXEval · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.8684, "skill": 0.7912, "coverage": 1, "random": 0.3697, "tracks": [], "combined": 0.6205, "private_eq": 0.7629}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 570

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · CRUXEval · Combined public/private component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.8684, "skill": 0.7912, "coverage": 1, "random": 0.3697, "tracks": [], "combined": 0.6205, "private_eq": 0.7629}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · CRUXEval · Equated private same-skill component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.8684, "skill": 0.7912, "coverage": 1, "random": 0.3697, "tracks": [], "combined": 0.6205, "private_eq": 0.7629}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · CLadder · accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.8542, "jev": null, "metric": "accuracy", "requests": 5000, "answered": 5000, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 44, "dataset": "CLadder", "short": "CLadder", "metric": "accuracy", "explainer": "Answers a yes/no question about cause and effect in a small made-up world with stated probabilities.", "baseline": {"value": 0.5, "label": "Uniform random choice"}, "lower": false, "range": null, "cases": 5000, "group": "knowledge", "in_suite": true, "interactive": false, "references": [{"kind": "paper", "url": "https://arxiv.org/abs/2312.04350"}, {"kind": "dataset", "url": "https://github.com/causalNLP/cladder"}], "private": true}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 5000

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · CLadder · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.8542, "skill": 0.7084, "coverage": 1, "random": 0.5, "tracks": [], "combined": 0.6638, "private_eq": 0.8333}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 5000

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · CLadder · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.8542, "skill": 0.7084, "coverage": 1, "random": 0.5, "tracks": [], "combined": 0.6638, "private_eq": 0.8333}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 5000

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · CLadder · Combined public/private component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.8542, "skill": 0.7084, "coverage": 1, "random": 0.5, "tracks": [], "combined": 0.6638, "private_eq": 0.8333}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · CLadder · Equated private same-skill component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.8542, "skill": 0.7084, "coverage": 1, "random": 0.5, "tracks": [], "combined": 0.6638, "private_eq": 0.8333}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · HLE · accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.1996, "jev": null, "metric": "accuracy", "requests": 501, "answered": 501, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 45, "dataset": "HLE", "short": "HLE", "metric": "accuracy", "explainer": "Expert-level multiple choice from Humanity's Last Exam, text-only questions, 5 to 26 options.", "baseline": {"value": 0.1641, "label": "Uniform random choice"}, "lower": false, "range": null, "cases": 501, "group": "knowledge", "in_suite": true, "interactive": false, "references": [{"kind": "dataset", "url": "https://huggingface.co/datasets/cais/hle"}], "gold": 1.2}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 501

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · HLE · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.1996, "skill": 0.0425, "coverage": 1, "random": 0.1641, "tracks": []}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 501

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · HLE · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.1996, "skill": 0.0425, "coverage": 1, "random": 0.1641, "tracks": []}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 501

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Habermas Machine · accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.5084, "jev": null, "metric": "accuracy", "requests": 1676, "answered": 1676, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 50, "dataset": "Habermas Machine", "short": "Habermas", "metric": "accuracy", "explainer": "Given a group's opinions on a policy question, predicts which candidate statement the group ranked highest.", "baseline": {"value": 0.311, "label": "Uniform random choice"}, "lower": false, "range": null, "cases": 1676, "group": "arts", "in_suite": true, "interactive": false, "references": [{"kind": "dataset", "url": "https://github.com/google-deepmind/habermas_machine"}], "private": true}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 1676

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Habermas Machine · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.5084, "skill": 0.2865, "coverage": 1, "random": 0.311, "tracks": [], "combined": 0.312, "private_eq": 0.5307}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 1676

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Habermas Machine · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.5084, "skill": 0.2865, "coverage": 1, "random": 0.311, "tracks": [], "combined": 0.312, "private_eq": 0.5307}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 1676

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Habermas Machine · Combined public/private component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.5084, "skill": 0.2865, "coverage": 1, "random": 0.311, "tracks": [], "combined": 0.312, "private_eq": 0.5307}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · Habermas Machine · Equated private same-skill component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.5084, "skill": 0.2865, "coverage": 1, "random": 0.311, "tracks": [], "combined": 0.312, "private_eq": 0.5307}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · PhishNChips phishing decisions · accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.752, "jev": null, "metric": "accuracy", "requests": 2000, "answered": 2000, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 56, "dataset": "PhishNChips phishing decisions", "short": "PhishNChips", "metric": "accuracy", "explainer": "Says whether an email and its link are phishing or legitimate.", "baseline": {"value": 0.5, "label": "uniform random choice, mean of 1/options"}, "lower": false, "range": null, "cases": 2000, "group": "retrieval", "in_suite": true, "interactive": false, "edition_added": "v0.2", "references": [{"kind": "dataset", "url": "https://huggingface.co/datasets/AreLit/PhishNChips"}, {"kind": "code", "url": "https://github.com/anisselbd/jev-phishing-bench"}], "private": true}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 2000

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · PhishNChips phishing decisions · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.752, "skill": 0.504, "coverage": 1, "random": 0.5, "tracks": [], "combined": 0.6203, "private_eq": 0.8118}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 2000

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · PhishNChips phishing decisions · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.752, "skill": 0.504, "coverage": 1, "random": 0.5, "tracks": [], "combined": 0.6203, "private_eq": 0.8118}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 2000

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · PhishNChips phishing decisions · Combined public/private component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.752, "skill": 0.504, "coverage": 1, "random": 0.5, "tracks": [], "combined": 0.6203, "private_eq": 0.8118}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · PhishNChips phishing decisions · Equated private same-skill component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.752, "skill": 0.504, "coverage": 1, "random": 0.5, "tracks": [], "combined": 0.6203, "private_eq": 0.8118}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · MMLU-Pro · accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.7283, "jev": null, "metric": "accuracy", "requests": 12032, "answered": 12032, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 57, "dataset": "MMLU-Pro", "short": "MMLU-Pro", "metric": "accuracy", "explainer": "Harder MMLU successor: questions across 14 subjects with up to ten options.", "baseline": {"value": 0.1109, "label": "uniform random choice, mean of 1/options"}, "lower": false, "range": null, "cases": 12032, "group": "knowledge", "in_suite": true, "interactive": false, "edition_added": "v0.2", "references": [{"kind": "paper", "url": "https://arxiv.org/abs/2406.01574"}, {"kind": "dataset", "url": "https://huggingface.co/datasets/TIGER-Lab/MMLU-Pro"}], "gold": 1.2, "private": true}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 12032

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · MMLU-Pro · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.7283, "skill": 0.6944, "coverage": 1, "random": 0.1109, "tracks": [], "private_eq": 0.6328, "combined": 0.5833}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 12032

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · MMLU-Pro · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.7283, "skill": 0.6944, "coverage": 1, "random": 0.1109, "tracks": [], "private_eq": 0.6328, "combined": 0.5833}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 12032

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · MMLU-Pro · Combined public/private component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.7283, "skill": 0.6944, "coverage": 1, "random": 0.1109, "tracks": [], "private_eq": 0.6328, "combined": 0.5833}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · MMLU-Pro · Equated private same-skill component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.7283, "skill": 0.6944, "coverage": 1, "random": 0.1109, "tracks": [], "private_eq": 0.6328, "combined": 0.5833}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · BBH · accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.9368, "jev": null, "metric": "accuracy", "requests": 5507, "answered": 5507, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 58, "dataset": "BBH", "short": "BBH", "metric": "accuracy", "explainer": "23 tricky reasoning tasks from BIG-Bench (logic, dates, tracking objects, fallacies), answered with no working.", "baseline": {"value": 0.3101, "label": "uniform random choice, mean of 1/options"}, "lower": false, "range": null, "cases": 5507, "group": "knowledge", "in_suite": true, "interactive": false, "edition_added": "v0.2", "references": [{"kind": "paper", "url": "https://arxiv.org/abs/2210.09261"}, {"kind": "dataset", "url": "https://github.com/suzgunmirac/BIG-Bench-Hard"}], "gold": 1.2, "private": true}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 5507

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · BBH · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.9368, "skill": 0.9084, "coverage": 1, "random": 0.3101, "tracks": [], "private_eq": 0.5295, "combined": 0.3112}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 5507

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · BBH · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.9368, "skill": 0.9084, "coverage": 1, "random": 0.3101, "tracks": [], "private_eq": 0.5295, "combined": 0.3112}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 5507

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · BBH · Combined public/private component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.9368, "skill": 0.9084, "coverage": 1, "random": 0.3101, "tracks": [], "private_eq": 0.5295, "combined": 0.3112}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · BBH · Equated private same-skill component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.9368, "skill": 0.9084, "coverage": 1, "random": 0.3101, "tracks": [], "private_eq": 0.5295, "combined": 0.3112}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · RAGTruth · F1 on hallucinated class

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.8483, "jev": null, "metric": "F1 on hallucinated class", "requests": 2700, "answered": 2700, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 59, "dataset": "RAGTruth", "short": "RAGTruth", "metric": "F1 on hallucinated class", "explainer": "Says yes or no: does an AI-written answer contain anything its source text doesn't support? Scored with F1.", "baseline": {"value": 0.5177, "label": "Always \"hallucinated\" · F1"}, "lower": false, "range": null, "cases": 2700, "group": "language", "in_suite": true, "interactive": false, "edition_added": "v0.2", "references": [{"kind": "dataset", "url": "https://github.com/ParticleMedia/RAGTruth"}], "private": true}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 2700

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · RAGTruth · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native F1 on hallucinated class. Source normalization: {"raw": 0.8483, "skill": 0.6855, "coverage": 1, "random": 0.5177, "tracks": [], "combined": 0.2959, "private_eq": 0.6638}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 2700

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · RAGTruth · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native F1 on hallucinated class. Source normalization: {"raw": 0.8483, "skill": 0.6855, "coverage": 1, "random": 0.5177, "tracks": [], "combined": 0.2959, "private_eq": 0.6638}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 2700

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · RAGTruth · Combined public/private component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native F1 on hallucinated class. Source normalization: {"raw": 0.8483, "skill": 0.6855, "coverage": 1, "random": 0.5177, "tracks": [], "combined": 0.2959, "private_eq": 0.6638}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · RAGTruth · Equated private same-skill component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native F1 on hallucinated class. Source normalization: {"raw": 0.8483, "skill": 0.6855, "coverage": 1, "random": 0.5177, "tracks": [], "combined": 0.2959, "private_eq": 0.6638}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · HoVer · accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.8788, "jev": null, "metric": "accuracy", "requests": 4000, "answered": 4000, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 61, "dataset": "HoVer", "short": "HoVer", "metric": "accuracy", "explainer": "Decides whether a claim linking two to four Wikipedia articles is supported by their introductions.", "baseline": {"value": 0.5, "label": "uniform random choice, mean of 1/options"}, "lower": false, "range": null, "cases": 4000, "group": "retrieval", "in_suite": true, "interactive": false, "edition_added": "v0.2", "references": [{"kind": "paper", "url": "https://arxiv.org/abs/2011.03088"}, {"kind": "dataset", "url": "https://github.com/hover-nlp/hover"}], "private": true}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 4000

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · HoVer · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.8788, "skill": 0.7576, "coverage": 1, "random": 0.5, "tracks": [], "combined": 0.6777, "private_eq": 0.8402}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 4000

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · HoVer · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.8788, "skill": 0.7576, "coverage": 1, "random": 0.5, "tracks": [], "combined": 0.6777, "private_eq": 0.8402}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 4000

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · HoVer · Combined public/private component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.8788, "skill": 0.7576, "coverage": 1, "random": 0.5, "tracks": [], "combined": 0.6777, "private_eq": 0.8402}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · HoVer · Equated private same-skill component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.8788, "skill": 0.7576, "coverage": 1, "random": 0.5, "tracks": [], "combined": 0.6777, "private_eq": 0.8402}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · When2Call · accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.8639, "jev": null, "metric": "accuracy", "requests": 3652, "answered": 3652, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 62, "dataset": "When2Call", "short": "When2Call", "metric": "accuracy", "explainer": "Given a question and an assistant's tools, picks the right move: answer, call a tool, ask for details, or decline.", "baseline": {"value": 0.25, "label": "uniform random choice, mean of 1/options"}, "lower": false, "range": null, "cases": 3652, "group": "tools", "in_suite": true, "interactive": false, "edition_added": "v0.2", "references": [{"kind": "paper", "url": "https://aclanthology.org/2025.naacl-long.174"}, {"kind": "dataset", "url": "https://huggingface.co/datasets/nvidia/When2Call"}], "private": true}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 3652

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · When2Call · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.8639, "skill": 0.8185, "coverage": 1, "random": 0.25, "tracks": [], "combined": 0.6725, "private_eq": 0.7564}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 3652

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · When2Call · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.8639, "skill": 0.8185, "coverage": 1, "random": 0.25, "tracks": [], "combined": 0.6725, "private_eq": 0.7564}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 3652

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · When2Call · Combined public/private component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.8639, "skill": 0.8185, "coverage": 1, "random": 0.25, "tracks": [], "combined": 0.6725, "private_eq": 0.7564}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · When2Call · Equated private same-skill component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.8639, "skill": 0.8185, "coverage": 1, "random": 0.25, "tracks": [], "combined": 0.6725, "private_eq": 0.7564}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · New Yorker · accuracy

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Native source result: {"score": 0.8049, "jev": null, "metric": "accuracy", "requests": 528, "answered": 528, "unsupported": 0, "errors": 0, "median_ms": null}. Benchmark contract: {"id": 64, "dataset": "New Yorker", "short": "New Yorker", "metric": "accuracy", "explainer": "Given a text description of a New Yorker cartoon, picks the caption written for it from five.", "baseline": {"value": 0.2, "label": "uniform random choice, mean of 1/options"}, "lower": false, "range": null, "cases": 528, "group": "arts", "in_suite": true, "interactive": false, "edition_added": "v0.2", "references": [{"kind": "paper", "url": "https://arxiv.org/abs/2209.06293"}, {"kind": "dataset", "url": "https://huggingface.co/datasets/jmhessel/newyorker_caption_contest"}], "private": true}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 528

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · New Yorker · Raw index component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.8049, "skill": 0.7561, "coverage": 1, "random": 0.2, "tracks": [], "combined": 0.7585, "private_eq": 0.8082}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 528

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · New Yorker · Chance-corrected public skill

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.8049, "skill": 0.7561, "coverage": 1, "random": 0.2, "tracks": [], "combined": 0.7585, "private_eq": 0.8082}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: 528

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · New Yorker · Combined public/private component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.8049, "skill": 0.7561, "coverage": 1, "random": 0.2, "tracks": [], "combined": 0.7585, "private_eq": 0.8082}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

HF text 0.4 · New Yorker · Equated private same-skill component

Exact source engine drex-max-v1; {"kind": "full fine-tune", "technique": "autoregressive", "technique_name": "autoregressive", "served_params": 26085330160, "params_estimated": false, "params_from_name": false, "trained_bytes": null, "model_url": null, "code_url": null, "base_model": "Qwen/Qwen3.8-27B", "base_url": "https://huggingface.co/Qwen/Qwen3.8-27B", "weights_repo": null, "served_checkpoint": null, "weights_note": "Open weights coming soon", "org": {"id": "Nace-DeepInfra", "name": "Nace-Deepinfra", "logo": "data/orgs/Nace-DeepInfra.png"}, "new": true, "pr": 238}. New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement.

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred. Source notes: New in the 2026-10-18 PR batch (PR #238). Submitted public 0.3 run adopted (kit rescore 67.52, equal to the claim). Private tests run by the lab on one RTX PRO 6000 with the author's serving (submitted endpoint undeclared, lab run bf16); decision agreement with the submitted rows 98.52% on 831 rows (bar 96%). A separate model from Drex v1.5 and Drex v2.0, which stay on the board. The submitted run used a private endpoint; the lab ran the published bf16 loader. Listed under a maintainer decision: one smoke row ran out of memory in the first check and matched the submitted answers once re-run. Per-benchmark times are not shown: the board run sent requests in batches, so its timings are batch wall-clock. The latency shown is the single-request measurement. Author public-run adoption/execution status and private lab setup are disclosed in configuration; evaluator-published does not mean every public run was independently executed. Normalized/index component, not native accuracy. Source normalization: {"raw": 0.8049, "skill": 0.7561, "coverage": 1, "random": 0.2, "tracks": [], "combined": 0.7585, "private_eq": 0.8082}

Decision Index 0.4 (release-v4), source generated 2026-10-20T14:13:56Z. Full = 60% equated private same-skill + 40% equated private new-domain tasks; public contributes zero. Both private components are mapped onto the public scale using untuned stock models. Public weighting/panel changed (GSM8K rebuilt, ForecastBench retired, WinoGrande area reassigned); these are new benchmark definitions, not proof of a model upgrade. Public 37-benchmark index and 42 published native benchmarks remain separate. Preserve source rounding, coverage, exclusions and unknown execution dates; no new uncertainty inferred.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-20T14:13:56Z · Revision: 133313cf8963f2b6bed960abb0152fd14cbb6731 · Published observations; result reuse terms not specified

Nace Drex Max launch · Decision Index (public)

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max launch · JevBench

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max launch · Decision Bench

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max launch · DecideBench (item)

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max launch · DecideBench (pair)

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max launch · Game Bench (win %)

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max launch · MMLU-Pro

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max launch · HLE

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max launch · CRUXEval

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max launch · ARC-Challenge

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max launch · ARC-Easy

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max launch · HellaSwag

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max launch · OSWorld-Verified (automation)

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max launch · OSWorld-V2 (automation)

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max launch · BFCL (tool calling)

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max launch · ToolRet (tool retrieval)

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max launch · When2Call

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max launch · RAGTruth

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max launch · Median API latency

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max launch · Cost (per 1k decisions)

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs. Workload-dependent cost per1000decisions, not an input-token tariff.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs. Workload-dependent cost per1000decisions, not an input-token tariff.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max launch vision · Public Vision Index

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs. Vendor public vision suite; unknown exact revision. Not the community 0.4 private/public composite.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs. Vendor public vision suite; unknown exact revision. Not the community 0.4 private/public composite.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max launch vision · CV-Bench

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs. Vendor public vision suite; unknown exact revision. Not the community 0.4 private/public composite.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs. Vendor public vision suite; unknown exact revision. Not the community 0.4 private/public composite.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max launch vision · BLINK

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs. Vendor public vision suite; unknown exact revision. Not the community 0.4 private/public composite.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs. Vendor public vision suite; unknown exact revision. Not the community 0.4 private/public composite.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max launch vision · RealWorldQA

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs. Vendor public vision suite; unknown exact revision. Not the community 0.4 private/public composite.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs. Vendor public vision suite; unknown exact revision. Not the community 0.4 private/public composite.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max launch vision · Winoground

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs. Vendor public vision suite; unknown exact revision. Not the community 0.4 private/public composite.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs. Vendor public vision suite; unknown exact revision. Not the community 0.4 private/public composite.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max launch vision · CharXiv

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs. Vendor public vision suite; unknown exact revision. Not the community 0.4 private/public composite.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs. Vendor public vision suite; unknown exact revision. Not the community 0.4 private/public composite.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max launch vision · InfographicVQA

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs. Vendor public vision suite; unknown exact revision. Not the community 0.4 private/public composite.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs. Vendor public vision suite; unknown exact revision. Not the community 0.4 private/public composite.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max launch vision · KIE (CORD + FUNSD)

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs. Vendor public vision suite; unknown exact revision. Not the community 0.4 private/public composite.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs. Vendor public vision suite; unknown exact revision. Not the community 0.4 private/public composite.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max launch vision · Mind2Web

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs. Vendor public vision suite; unknown exact revision. Not the community 0.4 private/public composite.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs. Vendor public vision suite; unknown exact revision. Not the community 0.4 private/public composite.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max launch vision · R-Bench-M

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs. Vendor public vision suite; unknown exact revision. Not the community 0.4 private/public composite.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs. Vendor public vision suite; unknown exact revision. Not the community 0.4 private/public composite.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max launch vision · MMMU-Pro vision

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs. Vendor public vision suite; unknown exact revision. Not the community 0.4 private/public composite.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs. Vendor public vision suite; unknown exact revision. Not the community 0.4 private/public composite.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max launch vision · Hateful Memes (moderation)

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs. Vendor public vision suite; unknown exact revision. Not the community 0.4 private/public composite.

Nace.AI 19 October 2026 release comparison; exact model columns, public native tables. Vendor has not specified all dataset revisions, sample populations, execution dates or reasoning/agent harnesses. Table labels and units retained; do not equate with independent/community runs. Vendor public vision suite; unknown exact revision. Not the community 0.4 private/public composite.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace vendor public0.3.2 · Public index

DeepInfra-published Nace vendor model-card figures; public Decision Index 0.3.2, 37 benchmarks/five areas, 140178 scored requests. Text/JSON-only provider; images rejected. Separate from community 110201-request submission. Unknown vendor run revision/date; no automatic first-place claim.

DeepInfra-published Nace vendor model-card figures; public Decision Index 0.3.2, 37 benchmarks/five areas, 140178 scored requests. Text/JSON-only provider; images rejected. Separate from community 110201-request submission. Unknown vendor run revision/date; no automatic first-place claim.

DeepInfra-published Nace vendor model-card figures; public Decision Index 0.3.2, 37 benchmarks/five areas, 140178 scored requests. Text/JSON-only provider; images rejected. Separate from community 110201-request submission. Unknown vendor run revision/date; no automatic first-place claim.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace vendor public0.3.2 · Knowledge and reasoning

DeepInfra-published Nace vendor model-card figures; public Decision Index 0.3.2, 37 benchmarks/five areas, 140178 scored requests. Text/JSON-only provider; images rejected. Separate from community 110201-request submission. Unknown vendor run revision/date; no automatic first-place claim.

DeepInfra-published Nace vendor model-card figures; public Decision Index 0.3.2, 37 benchmarks/five areas, 140178 scored requests. Text/JSON-only provider; images rejected. Separate from community 110201-request submission. Unknown vendor run revision/date; no automatic first-place claim.

DeepInfra-published Nace vendor model-card figures; public Decision Index 0.3.2, 37 benchmarks/five areas, 140178 scored requests. Text/JSON-only provider; images rejected. Separate from community 110201-request submission. Unknown vendor run revision/date; no automatic first-place claim.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace vendor public0.3.2 · Language

DeepInfra-published Nace vendor model-card figures; public Decision Index 0.3.2, 37 benchmarks/five areas, 140178 scored requests. Text/JSON-only provider; images rejected. Separate from community 110201-request submission. Unknown vendor run revision/date; no automatic first-place claim.

DeepInfra-published Nace vendor model-card figures; public Decision Index 0.3.2, 37 benchmarks/five areas, 140178 scored requests. Text/JSON-only provider; images rejected. Separate from community 110201-request submission. Unknown vendor run revision/date; no automatic first-place claim.

DeepInfra-published Nace vendor model-card figures; public Decision Index 0.3.2, 37 benchmarks/five areas, 140178 scored requests. Text/JSON-only provider; images rejected. Separate from community 110201-request submission. Unknown vendor run revision/date; no automatic first-place claim.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace vendor public0.3.2 · Retrieval

DeepInfra-published Nace vendor model-card figures; public Decision Index 0.3.2, 37 benchmarks/five areas, 140178 scored requests. Text/JSON-only provider; images rejected. Separate from community 110201-request submission. Unknown vendor run revision/date; no automatic first-place claim.

DeepInfra-published Nace vendor model-card figures; public Decision Index 0.3.2, 37 benchmarks/five areas, 140178 scored requests. Text/JSON-only provider; images rejected. Separate from community 110201-request submission. Unknown vendor run revision/date; no automatic first-place claim.

DeepInfra-published Nace vendor model-card figures; public Decision Index 0.3.2, 37 benchmarks/five areas, 140178 scored requests. Text/JSON-only provider; images rejected. Separate from community 110201-request submission. Unknown vendor run revision/date; no automatic first-place claim.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace vendor public0.3.2 · Tools

DeepInfra-published Nace vendor model-card figures; public Decision Index 0.3.2, 37 benchmarks/five areas, 140178 scored requests. Text/JSON-only provider; images rejected. Separate from community 110201-request submission. Unknown vendor run revision/date; no automatic first-place claim.

DeepInfra-published Nace vendor model-card figures; public Decision Index 0.3.2, 37 benchmarks/five areas, 140178 scored requests. Text/JSON-only provider; images rejected. Separate from community 110201-request submission. Unknown vendor run revision/date; no automatic first-place claim.

DeepInfra-published Nace vendor model-card figures; public Decision Index 0.3.2, 37 benchmarks/five areas, 140178 scored requests. Text/JSON-only provider; images rejected. Separate from community 110201-request submission. Unknown vendor run revision/date; no automatic first-place claim.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace vendor public0.3.2 · Arts

DeepInfra-published Nace vendor model-card figures; public Decision Index 0.3.2, 37 benchmarks/five areas, 140178 scored requests. Text/JSON-only provider; images rejected. Separate from community 110201-request submission. Unknown vendor run revision/date; no automatic first-place claim.

DeepInfra-published Nace vendor model-card figures; public Decision Index 0.3.2, 37 benchmarks/five areas, 140178 scored requests. Text/JSON-only provider; images rejected. Separate from community 110201-request submission. Unknown vendor run revision/date; no automatic first-place claim.

DeepInfra-published Nace vendor model-card figures; public Decision Index 0.3.2, 37 benchmarks/five areas, 140178 scored requests. Text/JSON-only provider; images rejected. Separate from community 110201-request submission. Unknown vendor run revision/date; no automatic first-place claim.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace vendor · JevBench 231 public items

DeepInfra-published Nace vendor model-card figures; public Decision Index 0.3.2, 37 benchmarks/five areas, 140178 scored requests. Text/JSON-only provider; images rejected. Separate from community 110201-request submission. Unknown vendor run revision/date; no automatic first-place claim.

DeepInfra-published Nace vendor model-card figures; public Decision Index 0.3.2, 37 benchmarks/five areas, 140178 scored requests. Text/JSON-only provider; images rejected. Separate from community 110201-request submission. Unknown vendor run revision/date; no automatic first-place claim. Exact native card table; do not substitute for Leanmcp/independent runs.

DeepInfra-published Nace vendor model-card figures; public Decision Index 0.3.2, 37 benchmarks/five areas, 140178 scored requests. Text/JSON-only provider; images rejected. Separate from community 110201-request submission. Unknown vendor run revision/date; no automatic first-place claim. Exact native card table; do not substitute for Leanmcp/independent runs.

Checked / reported: 2026-10-21 · Samples: 231

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace vendor · JevBench hard items

DeepInfra-published Nace vendor model-card figures; public Decision Index 0.3.2, 37 benchmarks/five areas, 140178 scored requests. Text/JSON-only provider; images rejected. Separate from community 110201-request submission. Unknown vendor run revision/date; no automatic first-place claim.

DeepInfra-published Nace vendor model-card figures; public Decision Index 0.3.2, 37 benchmarks/five areas, 140178 scored requests. Text/JSON-only provider; images rejected. Separate from community 110201-request submission. Unknown vendor run revision/date; no automatic first-place claim. Exact native card table; do not substitute for Leanmcp/independent runs.

DeepInfra-published Nace vendor model-card figures; public Decision Index 0.3.2, 37 benchmarks/five areas, 140178 scored requests. Text/JSON-only provider; images rejected. Separate from community 110201-request submission. Unknown vendor run revision/date; no automatic first-place claim. Exact native card table; do not substitute for Leanmcp/independent runs.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace vendor · Decision Bench v4

DeepInfra-published Nace vendor model-card figures; public Decision Index 0.3.2, 37 benchmarks/five areas, 140178 scored requests. Text/JSON-only provider; images rejected. Separate from community 110201-request submission. Unknown vendor run revision/date; no automatic first-place claim.

DeepInfra-published Nace vendor model-card figures; public Decision Index 0.3.2, 37 benchmarks/five areas, 140178 scored requests. Text/JSON-only provider; images rejected. Separate from community 110201-request submission. Unknown vendor run revision/date; no automatic first-place claim. Exact native card table; do not substitute for Leanmcp/independent runs.

DeepInfra-published Nace vendor model-card figures; public Decision Index 0.3.2, 37 benchmarks/five areas, 140178 scored requests. Text/JSON-only provider; images rejected. Separate from community 110201-request submission. Unknown vendor run revision/date; no automatic first-place claim. Exact native card table; do not substitute for Leanmcp/independent runs.

Checked / reported: 2026-10-21 · Samples: 1071

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max product · 243 screenshot question median API latency

Nace Drex product page checked 21 October 2026; exact Drex Max versus Jev 1.14 columns. Product benchmark population/configuration not fully specified, separate from community and release comparisons.

Nace Drex product page checked 21 October 2026; exact Drex Max versus Jev 1.14 columns. Product benchmark population/configuration not fully specified, separate from community and release comparisons. Same 243 screenshot questions, same client, one request at a time. Network-inclusive workload latency.

Nace Drex product page checked 21 October 2026; exact Drex Max versus Jev 1.14 columns. Product benchmark population/configuration not fully specified, separate from community and release comparisons. Same 243 screenshot questions, same client, one request at a time. Network-inclusive workload latency.

Checked / reported: 2026-10-21 · Samples: 243

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max product · FinEntity

Nace Drex product page checked 21 October 2026; exact Drex Max versus Jev 1.14 columns. Product benchmark population/configuration not fully specified, separate from community and release comparisons.

Nace Drex product page checked 21 October 2026; exact Drex Max versus Jev 1.14 columns. Product benchmark population/configuration not fully specified, separate from community and release comparisons. Vendor native public benchmark observation; do not blend with different submitted-run populations.

Nace Drex product page checked 21 October 2026; exact Drex Max versus Jev 1.14 columns. Product benchmark population/configuration not fully specified, separate from community and release comparisons. Vendor native public benchmark observation; do not blend with different submitted-run populations.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max product · BANKING77

Nace Drex product page checked 21 October 2026; exact Drex Max versus Jev 1.14 columns. Product benchmark population/configuration not fully specified, separate from community and release comparisons.

Nace Drex product page checked 21 October 2026; exact Drex Max versus Jev 1.14 columns. Product benchmark population/configuration not fully specified, separate from community and release comparisons. Vendor native public benchmark observation; do not blend with different submitted-run populations.

Nace Drex product page checked 21 October 2026; exact Drex Max versus Jev 1.14 columns. Product benchmark population/configuration not fully specified, separate from community and release comparisons. Vendor native public benchmark observation; do not blend with different submitted-run populations.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max product · ContractNLI

Nace Drex product page checked 21 October 2026; exact Drex Max versus Jev 1.14 columns. Product benchmark population/configuration not fully specified, separate from community and release comparisons.

Nace Drex product page checked 21 October 2026; exact Drex Max versus Jev 1.14 columns. Product benchmark population/configuration not fully specified, separate from community and release comparisons. Vendor native public benchmark observation; do not blend with different submitted-run populations.

Nace Drex product page checked 21 October 2026; exact Drex Max versus Jev 1.14 columns. Product benchmark population/configuration not fully specified, separate from community and release comparisons. Vendor native public benchmark observation; do not blend with different submitted-run populations.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified

Nace Drex Max product · Screenshot acceptance

Nace Drex product page checked 21 October 2026; exact Drex Max versus Jev 1.14 columns. Product benchmark population/configuration not fully specified, separate from community and release comparisons.

Nace Drex product page checked 21 October 2026; exact Drex Max versus Jev 1.14 columns. Product benchmark population/configuration not fully specified, separate from community and release comparisons. Input acceptance diagnostic, not answer accuracy. Jev rejects a side under280px or over1568px. Denominator not explicitly restated for this percentage; no inferred complement.

Nace Drex product page checked 21 October 2026; exact Drex Max versus Jev 1.14 columns. Product benchmark population/configuration not fully specified, separate from community and release comparisons. Input acceptance diagnostic, not answer accuracy. Jev rejects a side under280px or over1568px. Denominator not explicitly restated for this percentage; no inferred complement.

Checked / reported: 2026-10-21 · Samples: Not published

Snapshot: 2026-10-19 · Revision: 2026-10-21 · Published observations; result reuse terms not specified