Mathematical Reasoning on MATH (Accuracy (%))
90.8Accuracy (%)Pass@k
Evaluation Results
| Method | Links | |
|---|---|---|
| Pass@kk=16, Model=Qwen2.5-7B2026.06 | 90.8 | |
| PRM-72B guided searchk=16, L=20, Base Model=Qwen2.5-7B, Guidance Model=Qwen2.5-72B2026.06 | 84 | |
| CGSk=16, L=20, Base Model=Qwen2.5-7B, Guidance Model=Qwen2.5-72B2026.06 | 81.8 | |
| Large greedyModel=Qwen2.5-72B2026.06 | 80.6 | |
| LGSk=16, L=20, Base Model=Qwen2.5-7B, Guidance Model=Qwen2.5-72B2026.06 | 80.2 | |
| Borda count (small)k=16, Model=Qwen2.5-7B2026.06 | 78 | |
| Borda count (large)k=16, Model=Qwen2.5-72B2026.06 | 77.4 | |
| Majority@kk=16, Model=Qwen2.5-7B2026.06 | 77.2 | |
| Self-Certainty (large)k=16, Model=Qwen2.5-72B2026.06 | 75.8 | |
| Best-of-Nk=16, Model=Qwen2.5-7B2026.06 | 75.8 | |
| Self-Certainty (small)k=16, Model=Qwen2.5-7B2026.06 | 72.8 | |
| Small greedyModel=Qwen2.5-7B2026.06 | 70 |