Overrefusal Evaluation on SQL-1k
0.3Refusal Rate (RR)Baseline
Evaluation Results
| Method | Links | |
|---|---|---|
| BaselineSetting / Model=P-SFT, Llama22026.03 | 0.3 | |
| BaselineSetting / Model=RLVR, Llama3-U2026.03 | 0.8 | |
| Db as Our DataSetting / Model=RLVR, Llama3-U2026.03 | 1.3 | |
| Db as AlpacaSetting / Model=RLVR, Llama3-U2026.03 | 2.1 | |
| Db as Our DataSetting / Model=P-SFT, Llama22026.03 | 3.4 | |
| Db as AlpacaSetting / Model=P-SFT, Llama22026.03 | 10.2 |