| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Backdoor Attack | DAN (Do-Anything-Now) | ASRw88.07 | 48 | |
| Safety Evaluation | DAN | Safety Score (DAN)91 | 26 | |
| Harmful Refusal | DAN | ASR94.9 | 16 | |
| Jailbreak Defense | DAN | Drop in ASR42.9 | 6 | |
| Jailbreak Detection | DAN | Detection Rate100 | 4 | |
| Jailbreak Robustness | DAN Static | ASR74.8 | 3 |