Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Jailbreak Robustness on AutoDAN Harm single-turn attack
Loading...
76.9
Attack Success Rate (ASR)
No Defense
-3.076
17.687
38.45
59.213
Jun 1, 2026
Attack Success Rate (ASR)
Updated 1d ago
Evaluation Results
Method
Method
Links
Attack Success Rate (ASR)
No Defense
Model=Qwen2.5-7B
2026.06
76.9
No Defense
Model=Llama3-8B
2026.06
60
PROACT
Model=Qwen2.5-7B
2026.06
0
SAGE
Model=Qwen2.5-7B
2026.06
0
THRD
Model=Qwen2.5-7B
2026.06
0
PROACT
Model=Llama3-8B
2026.06
0
SAGE
Model=Llama3-8B
2026.06
0
THRD
Model=Llama3-8B
2026.06
0
Feedback
Search any
task
Search any
task