Jailbreak Attack Robustness on FigStep jailbreak attack
17.44DSRTGA-7B
Evaluation Results
| Method | Links | |
|---|---|---|
| TGA-7Bparameters=7B2024.10 | 17.44 | |
| Unlearn-FigS2024.10 | 10.57 | |
| Qwen-VL-chatvariant=chat2024.10 | 2.91 | |
| InstructBlip2024.10 | 2.47 | |
| LLaVA-v1.6version=v1.62024.10 | 0.6 | |
| LLaVA-v1.5version=v1.52024.10 | 0.58 | |
| BLIP-22024.10 | 0 |