Mathematical Reasoning on MATH500 OOD (Pass@1, #Tokens)
93Pass@1 AccuracyDRP
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| DRPBackbone=R1-Distill-Qwen-7B, Evaluation Setting=zero-shot2025.05 | 93 | 1,781 | |
| BaseBackbone=R1-Distill-Qwen-7B, Evaluation Setting=zero-shot2025.05 | 92.4 | 2,486 | |
| TALEBackbone=R1-Distill-Qwen-7B, Evaluation Setting=zero-shot2025.05 | 91.6 | 2,530 | |
| Cot ValveBackbone=R1-Distill-Qwen-7B, Evaluation Setting=zero-shot2025.05 | 89.4 | 1,975 | |
| DRPBackbone=R1-Distill-Qwen-1.5B, Evaluation Setting=zero-shot2025.05 | 82 | 2,122 | |
| BaseBackbone=R1-Distill-Qwen-1.5B, Evaluation Setting=zero-shot2025.05 | 80.4 | 3,276 | |
| ThinkpruneBackbone=R1-Distill-Qwen-1.5B, Evaluation Setting=zero-shot2025.05 | 79.2 | 2,006 | |
| Cot ValveBackbone=R1-Distill-Qwen-1.5B, Evaluation Setting=zero-shot2025.05 | 76.5 | 2,705 | |
| TALEBackbone=R1-Distill-Qwen-1.5B, Evaluation Setting=zero-shot2025.05 | 76.2 | 3,107 |