Safety Evaluation on JBench-H
0ASRDb as Alpaca
Evaluation Results
| Method | Links | |
|---|---|---|
| Db as AlpacaSetting / Model=P-SFT, Llama2, ASR evaluation protocol=prefill attacks2026.03 | 0 | |
| Db as AlpacaSetting / Model=RLVR, Llama3-U2026.03 | 0 | |
| Db as Our DataSetting / Model=P-SFT, Llama2, ASR evaluation protocol=prefill attacks2026.03 | 5 | |
| Db as Our DataSetting / Model=RLVR, Llama3-U2026.03 | 5 | |
| BaselineSetting / Model=P-SFT, Llama2, ASR evaluation protocol=prefill attacks2026.03 | 36 | |
| BaselineSetting / Model=RLVR, Llama3-U2026.03 | 80 |