Commonsense Question Answering on CSQA out-of-domain
49.3AccuracySuCo
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| SuCo2026.06 | 49.3 | 369 | |
| DeepSeek-R1-Distill2026.06 | 45 | 743 | |
| MFT2026.06 | 26.6 | 342 | |
| Full CoT SFT2026.06 | 19.4 | 1,061 |
| Method | Links | ||
|---|---|---|---|
| SuCo2026.06 | 49.3 | 369 | |
| DeepSeek-R1-Distill2026.06 | 45 | 743 | |
| MFT2026.06 | 26.6 | 342 | |
| Full CoT SFT2026.06 | 19.4 | 1,061 |