Math Reasoning on OlympiadB
90.7AccuracyAPRM
Evaluation Results
| Method | Links | |
|---|---|---|
| APRMBackbone=Gemma-3-27B2025.11 | 90.7 | |
| APRMBackbone=GPT-OSS-20B2025.11 | 90.4 | |
| APRMBackbone=GPT-OSS-120B2025.11 | 89.4 | |
| ToTBackbone=GPT-OSS-20B2025.11 | 85.6 | |
| ToTBackbone=Gemma-3-27B2025.11 | 85.2 | |
| ReST-MCTSBackbone=GPT-OSS-20B2025.11 | 84.7 | |
| ToTBackbone=GPT-OSS-120B2025.11 | 84.3 | |
| rStar MathBackbone=GPT-OSS-120B2025.11 | 84.3 | |
| ReST-MCTSBackbone=GPT-OSS-120B2025.11 | 84.3 | |
| rStar MathBackbone=GPT-OSS-20B2025.11 | 84.3 | |
| ReST-MCTSBackbone=Gemma-3-27B2025.11 | 84.3 | |
| AutoPSVBackbone=GPT-OSS-120B2025.11 | 84 | |
| CoT-SCBackbone=GPT-OSS-120B2025.11 | 83.9 | |
| LLM-JBackbone=GPT-OSS-20B2025.11 | 83.9 | |
| Math S.Backbone=GPT-OSS-120B2025.11 | 83.8 | |
| rStar MathBackbone=Gemma-3-27B2025.11 | 83.8 | |
| Verify SbS.Backbone=GPT-OSS-120B2025.11 | 83.5 | |
| LLM-JBackbone=GPT-OSS-120B2025.11 | 83.1 | |
| LLM-JBackbone=Gemma-3-27B2025.11 | 83.1 | |
| AutoPSVBackbone=GPT-OSS-20B2025.11 | 75 | |
| Math S.Backbone=GPT-OSS-20B2025.11 | 73 | |
| Verify SbS.Backbone=GPT-OSS-20B2025.11 | 70 | |
| AutoPSVBackbone=Gemma-3-27B2025.11 | 70 | |
| Math S.Backbone=Gemma-3-27B2025.11 | 65 | |
| CoT-SCBackbone=GPT-OSS-20B2025.11 | 61 | |
| Verify SbS.Backbone=Gemma-3-27B2025.11 | 60 | |
| CoT-SCBackbone=Gemma-3-27B2025.11 | 48.7 | |
| APRMBackbone=Gemma-3-12B2025.11 | 44 | |
| ToTBackbone=Gemma-3-12B2025.11 | 40.7 | |
| ReST-MCTSBackbone=Gemma-3-12B2025.11 | 40.3 | |
| rStar MathBackbone=Gemma-3-12B2025.11 | 39 | |
| AutoPSVBackbone=Gemma-3-12B2025.11 | 38 | |
| LLM-JBackbone=Gemma-3-12B2025.11 | 37.7 | |
| Math S.Backbone=Gemma-3-12B2025.11 | 37 | |
| Verify SbS.Backbone=Gemma-3-12B2025.11 | 36 | |
| CoT-SCBackbone=Gemma-3-12B2025.11 | 35.6 |