Jailbreak Attack on Harmful Prompts Victim: GPT-4o
4Runtime (H:MM)PAPILLON
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| PAPILLONVictim Model=GPT-4o2026.04 | 4 | 77 | 76 | |
| JAIL-CON-CITVictim Model=GPT-4o2026.04 | 6 | 3 | 7 | |
| JAIL-CON-CVTVictim Model=GPT-4o2026.04 | 7 | 0 | 1 | |
| EvoJailVictim Model=GPT-4o2026.04 | 8 | 93 | 97 | |
| AutoDANVictim Model=GPT-4o2026.04 | 41 | 19 | 31 |