Mathematical Problem Solving on AIME 2024 (Pass@12, Avg@12, Maj@12)
87.78Pass@12PW-OPSD Aggressive
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| PW-OPSD AggressiveModel=Qwen3-4B, Max Generation Length=38,912 tokens, Number of Evaluation Seeds=3, Schedule Type=Aggressive2026.05 | 87.78 | 75.19 | 80 | |
| OPSDModel=Qwen3-4B, Max Generation Length=38,912 tokens, Number of Evaluation Seeds=32026.05 | 86.67 | 75.19 | 80 | |
| REOPOLDModel=Qwen3-4B, Max Generation Length=38,912 tokens, Number of Evaluation Seeds=32026.05 | 86.67 | 73.98 | 80 | |
| EOPDModel=Qwen3-4B, Max Generation Length=38,912 tokens, Number of Evaluation Seeds=32026.05 | 85.56 | 73.61 | 80 | |
| PW-OPSD ModerateModel=Qwen3-4B, Max Generation Length=38,912 tokens, Number of Evaluation Seeds=3, Schedule Type=Moderate2026.05 | 85.56 | 76.2 | 80 |