Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Jailbreak Attack on MaliciousInstruct (test)
Loading...
95
ASR (Refusal)
DSN
22.2
41.1
60
78.9
Apr 25, 2024
ASR (Refusal)
ASR (HarmBench)
Updated 1mo ago
Evaluation Results
Method
Method
Links
ASR (Refusal)
ASR (HarmBench)
DSN
Model=Gemma2-9B
2024.04
95
93
GCG
Model=Gemma2-9B
2024.04
88
88
DSN
Model=Qwen2-7B
2024.04
84
88
GCG
Model=Llama-3.1-8B
2024.04
77
32
DSN
Model=Llama-3.1-8B
2024.04
77
47
DSN
Model=Llama-3-8B
2024.04
70
69
GCG
Model=Qwen2-7B
2024.04
54
71
DSN
Model=Llama-2-13B
2024.04
36
72
GCG
Model=Llama-3-8B
2024.04
29
34
GCG
Model=Llama-2-13B
2024.04
25
51
Feedback
Search any
task
Search any
task