Domain Reasoning on DRBench (test)
42.9ScoreDR-Rubric-14B (BS-2)
Evaluation Results
| Method | Links | |
|---|---|---|
| DR-Rubric-14B (BS-2)Model Scale=14B2026.05 | 42.9 | |
| DR-Rubric-30B-A3B (BS-2)Model Scale=30B-A3B2026.05 | 42.2 | |
| DR-Rubric-30B-A3B (BS-3)Model Scale=30B-A3B2026.05 | 42.1 | |
| Tongyi-DeepResearch-30B-A3BModel Scale=30B-A3B2026.05 | 41.9 | |
| DeepSeek-R1-Distill-Qwen-14BModel Scale=14B2026.05 | 41 | |
| DR-Rubric-14B (BS-1)Model Scale=14B2026.05 | 40.4 | |
| DR-Rubric-30B-A3B (BS-1)Model Scale=30B-A3B2026.05 | 40.1 | |
| Qwen3-14B-baseModel Scale=14B2026.05 | 38.7 | |
| MiroThinker-1.7-mini (30B-A3B)Model Scale=30B-A3B2026.05 | 38.7 | |
| DR-Rubric-14B (BS-3)Model Scale=14B2026.05 | 37.7 | |
| WebThinker-R1-14BModel Scale=14B2026.05 | 37.6 | |
| Qwen3-30B-A3BModel Scale=30B-A3B2026.05 | 37.5 | |
| WebThinker-32B-DPOModel Scale=30B-A3B2026.05 | 37.5 | |
| Ministral-3-14B-Reasoning-2512Model Scale=14B2026.05 | 34.9 |