Mathematical Problem Solving on MATH 500 (Pass@12, Avg@12, Maj@12)
98.4Pass@12REOPOLD
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| REOPOLDModel=Qwen3-4B, Max Generation Length=38,912 tokens, Number of Evaluation Seeds=32026.05 | 98.4 | 95.09 | 96.53 | |
| PW-OPSD ModerateModel=Qwen3-4B, Max Generation Length=38,912 tokens, Number of Evaluation Seeds=3, Schedule Type=Moderate2026.05 | 98.4 | 95.34 | 96.67 | |
| PW-OPSD AggressiveModel=Qwen3-4B, Max Generation Length=38,912 tokens, Number of Evaluation Seeds=3, Schedule Type=Aggressive2026.05 | 98.4 | 95.53 | 97.07 | |
| EOPDModel=Qwen3-4B, Max Generation Length=38,912 tokens, Number of Evaluation Seeds=32026.05 | 98.33 | 95.33 | 96.73 | |
| OPSDModel=Qwen3-4B, Max Generation Length=38,912 tokens, Number of Evaluation Seeds=32026.05 | 98 | 95.33 | 96.73 |