Mathematical Reasoning on MATH500 (ACC, LEN, NAG)
92.1Accuracy (ACC)CEEH-ME
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| CEEH-MEExecution Strategy=Online, Base Model=R1-Distill-Qwen2.5-7B2026.02 | 92.1 | 2,170 | -0.56 | |
| CEEH-EAExecution Strategy=Online, Base Model=R1-Distill-Qwen2.5-7B2026.02 | 91.7 | 2,327 | -0.27 | |
| Length-PenaltyExecution Strategy=Online, Base Model=R1-Distill-Qwen2.5-7B, Reproduced=true2026.02 | 91.4 | 2,696 | -0.06 | |
| R1-Distill-Qwen2.5-7BReproduced=true2026.02 | 91.3 | 3,701 | — | |
| ThinkSwitcherExecution Strategy=Offline, Base Model=R1-Distill-Qwen2.5-7B2026.02 | 91.3 | 3,495 | 0 | |
| AutoThinkExecution Strategy=Online, Base Model=R1-Distill-Qwen2.5-7B2026.02 | 91.2 | 2,146 | 0.07 | |
| ConCISE-SimPOExecution Strategy=Online, Base Model=R1-Distill-Qwen2.5-7B2026.02 | 91 | 1,945 | 0.23 | |
| SpiritExecution Strategy=Online, Base Model=R1-Distill-Qwen2.5-7B2026.02 | 90.8 | 1,765 | 0.4 | |
| LC-R1Execution Strategy=Online, Base Model=R1-Distill-Qwen2.5-7B2026.02 | 90.4 | 1,568 | 0.75 | |
| DEERExecution Strategy=Offline, Base Model=R1-Distill-Qwen2.5-7B2026.02 | 89.8 | 2,143 | 1.07 | |
| DASTExecution Strategy=Online, Base Model=R1-Distill-Qwen2.5-7B2026.02 | 89.6 | 2,162 | 1.2 | |
| Dynasor-CoTExecution Strategy=Offline, Base Model=R1-Distill-Qwen2.5-7B2026.02 | 89.4 | 2,661 | 1.1 | |
| AutoThinkExecution Strategy=Online, Base Model=R1-Distill-Qwen2.5-1.5B2026.02 | 84 | 2,195 | -2.25 | |
| CEEH-MEExecution Strategy=Online, Base Model=R1-Distill-Qwen2.5-1.5B2026.02 | 83.8 | 3,067 | -1.69 | |
| CEEH-EAExecution Strategy=Online, Base Model=R1-Distill-Qwen2.5-1.5B2026.02 | 83.6 | 2,497 | -1.78 | |
| R1-Distill-Qwen2.5-1.5BReproduced=true2026.02 | 81.5 | 4,772 | — | |
| Qwen2.5-7B-Math-InsExecution Strategy=Prompting2026.02 | 81.4 | 670 | 9.81 | |
| LC-R1Execution Strategy=Online, Base Model=R1-Distill-Qwen2.5-1.5B2026.02 | 79.3 | 1,822 | 2.12 | |
| Qwen2.5-7B-InsExecution Strategy=Prompting2026.02 | 74.2 | 567 | 17.24 | |
| Qwen2.5-7B-MathExecution Strategy=Prompting2026.02 | 63.4 | 740 | 27.33 |