Math on AIME 2025 (Top-1 score)
28.83Top-1 ScoreePF
Evaluation Results
| Method | Links | |
|---|---|---|
| ePFBackbone=Qwen2.5-7B-In, Weighting scheme=w2025.10 | 28.83 | |
| ePFBackbone=Qwen2.5-7B-In, Weighting scheme=u2025.10 | 25.1 | |
| ePFBackbone=Qwen2.5-7B-In, Weighting scheme=iw2025.10 | 21.96 | |
| PFBackbone=Qwen2.5-7B-In, Weighting scheme=w2025.10 | 21.61 | |
| PFBackbone=Qwen2.5-7B-In, Weighting scheme=u2025.10 | 19.8 | |
| PFBackbone=Qwen2.5-7B-In, Weighting scheme=iw2025.10 | 17.61 | |
| Best-of-NBackbone=Qwen2.5-7B-In, Weighting scheme=w2025.10 | 17.41 | |
| Beam-SearchBackbone=Qwen2.5-7B-In, Weighting scheme=iw2025.10 | 16.81 | |
| Beam-SearchBackbone=Qwen2.5-7B-In, Weighting scheme=u2025.10 | 16.2 | |
| Best-of-NBackbone=Qwen2.5-7B-In, Weighting scheme=u2025.10 | 15.8 | |
| Best-of-NBackbone=Qwen2.5-7B-In, Weighting scheme=iw2025.10 | 14.9 | |
| Beam-SearchBackbone=Qwen2.5-7B-In, Weighting scheme=w2025.10 | 14.19 | |
| ePFBackbone=Qwen2.5-1.5B-In, Weighting scheme=w2025.10 | 10.82 | |
| Beam-SearchBackbone=Qwen2.5-1.5B-In, Weighting scheme=w2025.10 | 9.45 | |
| Beam-SearchBackbone=Qwen2.5-1.5B-In, Weighting scheme=u2025.10 | 7.4 | |
| PFBackbone=Qwen2.5-1.5B-In, Weighting scheme=w2025.10 | 7.32 | |
| ePFBackbone=Qwen2.5-1.5B-In, Weighting scheme=u2025.10 | 7.28 | |
| Base SamplingBackbone=Qwen2.5-7B-In, Weighting scheme=u2025.10 | 6.66 | |
| Best-of-NBackbone=Qwen2.5-1.5B-In, Weighting scheme=w2025.10 | 5.13 | |
| PFBackbone=Qwen2.5-1.5B-In, Weighting scheme=u2025.10 | 4.5 | |
| Beam-SearchBackbone=Qwen2.5-1.5B-In, Weighting scheme=iw2025.10 | 4.32 | |
| Best-of-NBackbone=Qwen2.5-1.5B-In, Weighting scheme=u2025.10 | 3.6 | |
| ePFBackbone=Qwen2.5-1.5B-In, Weighting scheme=iw2025.10 | 3.42 | |
| Base SamplingBackbone=Qwen2.5-1.5B-In, Weighting scheme=u2025.10 | 3.33 | |
| PFBackbone=Qwen2.5-1.5B-In, Weighting scheme=iw2025.10 | 2.87 | |
| Best-of-NBackbone=Qwen2.5-1.5B-In, Weighting scheme=iw2025.10 | 2.52 |