Mathematical Reasoning on AIME 25 (Accuracy (pass@1))
37.29Accuracy (pass@1)Critique
Evaluation Results
| Method | Links | |
|---|---|---|
| CritiqueBackbone Model=DS-R1-Qwen-7B, Cmax=3, Bmax=92162025.06 | 37.29 | |
| WaitBackbone Model=DS-R1-Qwen-7B, Cmax=3, Bmax=92162025.06 | 36.77 | |
| AlternativelyBackbone Model=DS-R1-Qwen-7B, Cmax=3, Bmax=92162025.06 | 36.51 | |
| HmmBackbone Model=DS-R1-Qwen-7B, Cmax=3, Bmax=92162025.06 | 36.51 | |
| Baseline (w/o BF)Backbone Model=DS-R1-Qwen-7B, Cmax=3, Bmax=92162025.06 | 36.41 | |
| LearnedBackbone Model=DS-R1-Qwen-7B, Cmax=3, Bmax=92162025.06 | 36.09 | |
| WaitBackbone Model=DS-R1-Llama-8B, Cmax=3, Bmax=92162025.06 | 30.26 | |
| LearnedBackbone Model=DS-R1-Llama-8B, Cmax=3, Bmax=92162025.06 | 30.26 | |
| HmmBackbone Model=DS-R1-Llama-8B, Cmax=3, Bmax=92162025.06 | 30.1 | |
| CritiqueBackbone Model=DS-R1-Llama-8B, Cmax=3, Bmax=92162025.06 | 29.9 | |
| AlternativelyBackbone Model=DS-R1-Llama-8B, Cmax=3, Bmax=92162025.06 | 29.64 | |
| Baseline (w/o BF)Backbone Model=DS-R1-Llama-8B, Cmax=3, Bmax=92162025.06 | 29.32 | |
| HmmBackbone Model=DS-R1-Qwen-1.5B, Cmax=3, Bmax=92162025.06 | 23.85 | |
| AlternativelyBackbone Model=DS-R1-Qwen-1.5B, Cmax=3, Bmax=92162025.06 | 23.8 | |
| WaitBackbone Model=DS-R1-Qwen-1.5B, Cmax=3, Bmax=92162025.06 | 23.49 | |
| Baseline (w/o BF)Backbone Model=DS-R1-Qwen-1.5B, Cmax=3, Bmax=92162025.06 | 23.33 | |
| CritiqueBackbone Model=DS-R1-Qwen-1.5B, Cmax=3, Bmax=92162025.06 | 22.45 | |
| LearnedBackbone Model=DS-R1-Qwen-1.5B, Cmax=3, Bmax=92162025.06 | 22.4 |