Math Reasoning on AIME 2024 (Pass@1, Pass@16)
41.66Pass@1EAPO
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| EAPOModel=Qwen-3-8B-Base2026.06 | 41.66 | 74.37 | |
| EAPOModel=Qwen-2.5-Math-7B2026.06 | 40.14 | 60.44 | |
| EAPO w/o sISModel=Qwen-2.5-Math-7B2026.06 | 39.68 | 63.29 | |
| MOPDModel=Qwen-2.5-Math-7B2026.06 | 38.85 | 60.28 | |
| OPDModel=Qwen-3-8B-Base2026.06 | 38.23 | 67.48 | |
| OPDModel=Qwen-2.5-Math-7B2026.06 | 38.12 | 55.06 | |
| MOPDModel=Qwen-3-8B-Base2026.06 | 37.8 | 74.68 | |
| Trajectory ReplayModel=Qwen-2.5-Math-7B2026.06 | 37.25 | 58.6 | |
| DAPOModel=Qwen-2.5-Math-7B2026.06 | 37.19 | 58 | |
| DAPOModel=Qwen-3-8B-Base2026.06 | 36.67 | 71.41 | |
| GRPOModel=Qwen-2.5-Math-7B2026.06 | 32.08 | 50.24 | |
| GRPOModel=Qwen-3-8B-Base2026.06 | 31.67 | 61.98 |