Jailbreak Attack on ActorAttack (test)
54ASRClaude Sonnet 4.5
Evaluation Results
| Method | Links | |
|---|---|---|
| Claude Sonnet 4.5Safety Steering=Disabled2025.02 | 54 | |
| GPT-5Safety Steering=Disabled2025.02 | 35.5 | |
| Claude Sonnet 4.5Safety Steering=Enabled2025.02 | 11 | |
| GPT-5Safety Steering=Enabled2025.02 | 4 |