Safety Filter Bypass on Misbinding Prompt
44.6NSFW-TC ScoreAttribute Misbinding Attack
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| Attribute Misbinding AttackSource=-2025.12 | 44.6 | 73.13 | 99.8 | 51.59 | 46.37 | 40.56 |
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| Attribute Misbinding AttackSource=-2025.12 | 44.6 | 73.13 | 99.8 | 51.59 | 46.37 | 40.56 |