Scientific Reasoning on GPQA Diamond (Avg@64)
39.6Avg@64DARS-HW-Breadth
Evaluation Results
| Method | Links | |
|---|---|---|
| DARS-HW-BreadthBase Model=Qwen2.5-Math-7B2025.08 | 39.6 | |
| DARS-HWBase Model=Qwen2.5-Math-7B2025.08 | 38.6 | |
| RLVR BaselineBase Model=Qwen2.5-Math-7B2025.08 | 38.5 | |
| DARS-HW-BreadthBase Model=Qwen2.5-Math-1.5B2025.08 | 29.3 | |
| RLVR BaselineBase Model=Qwen2.5-Math-1.5B2025.08 | 28.5 | |
| DARS-HWBase Model=Qwen2.5-Math-1.5B2025.08 | 28.4 |