Code Generation on MBPP (Response Accuracy)
65.7Response AccuracySuCo
Evaluation Results
| Method | Links | |
|---|---|---|
| SuCoModel Scale=7B2026.06 | 65.7 | |
| LHRMsModel Scale=7B2026.06 | 62.4 | |
| S-GRPOModel Scale=7B2026.06 | 62.3 | |
| AdaptThinkModel Scale=7B2026.06 | 62 | |
| AdaCoTModel Scale=7B2026.06 | 61.7 | |
| DeepSeek-R1-DistillModel Scale=7B2026.06 | 57.6 | |
| SuCoModel Scale=1.5B2026.06 | 48.5 | |
| AdaCoTModel Scale=1.5B2026.06 | 44.3 | |
| LHRMsModel Scale=1.5B2026.06 | 43.1 | |
| S-GRPOModel Scale=1.5B2026.06 | 42.9 | |
| AdaptThinkModel Scale=1.5B2026.06 | 42.1 | |
| DeepSeek-R1-DistillModel Scale=1.5B2026.06 | 41 | |
| Math-InstructModel Scale=7B2026.06 | 26.4 | |
| Math-BaseModel Scale=7B2026.06 | 13.1 | |
| Math-InstructModel Scale=1.5B2026.06 | 6.1 | |
| Math-BaseModel Scale=1.5B2026.06 | 4 |