Mathematical Reasoning on AIME24 (mean@8)
39.79Mean@8 ScoreACPO
Evaluation Results
| Method | Links | |
|---|---|---|
| ACPOModel=DeepSeek-R1-Distill-Qwen-7B2026.07 | 39.79 | |
| SAPOModel=DeepSeek-R1-Distill-Qwen-7B2026.07 | 38.54 | |
| ACPOModel=Qwen2.5-Math-7B2026.07 | 36.56 | |
| GTPOModel=DeepSeek-R1-Distill-Qwen-7B2026.07 | 36.37 | |
| GTPOModel=Qwen2.5-Math-7B2026.07 | 33.97 | |
| SAPOModel=Qwen2.5-Math-7B2026.07 | 32.29 | |
| 80/20Model=DeepSeek-R1-Distill-Qwen-7B2026.07 | 32.16 | |
| ACPOModel=Qwen3-8B-Base2026.07 | 30.31 | |
| 80/20Model=Qwen2.5-Math-7B2026.07 | 29.25 | |
| SAPOModel=Qwen3-8B-Base2026.07 | 29.06 | |
| DAPOModel=DeepSeek-R1-Distill-Qwen-7B2026.07 | 29.06 | |
| GRPOModel=DeepSeek-R1-Distill-Qwen-7B2026.07 | 28.43 | |
| DAPOModel=Qwen2.5-Math-7B2026.07 | 27.7 | |
| GTPOModel=Qwen3-8B-Base2026.07 | 27.21 | |
| GRPOModel=Qwen2.5-Math-7B2026.07 | 26.25 | |
| 80/20Model=Qwen3-8B-Base2026.07 | 26.02 | |
| DAPOModel=Qwen3-8B-Base2026.07 | 24.89 | |
| GRPOModel=Qwen3-8B-Base2026.07 | 22.6 |