Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

JailbreakBench

Benchmarks

Task NameDataset NameSOTA ResultTrend
Jailbreak AttackJailbreakBench
ASR100
242
Jailbreak AttackJailbreakBench
ASR@100
132
Jailbreak RobustnessJailbreakBench
Harmbench ASR0
88
Jailbreak AttackJailbreakBench
Attack Success Rate (ASR)0
76
Unsafe RobustnessJailbreakBench
Unsafe Rate0
72
Jailbreak AttackJailbreakBench (JBB)
ASR0
62
Safety EvaluationJailbreakBench (JBB) (test)
ASR (Llama-Guard-3-8B)1.12
56
Refusal suppressionJailbreakBench (test)
Attack Success Rate (ASR)0
54
JailbreakingJailbreakBench
Attack Success Rate (ASR)2
53
JailbreakingJailbreakBench
ASR85
42
Jailbreak AttackJailbreakBench (JBB) (test)
Attack Success Rate (ASR)98
42
Jailbreak AttackJailbreakBench
Attack Success Rate (ASR)96
40
Jailbreak AttackJailbreakBench
ASR197
39
JailbreakJailbreakBench (original split)
ASR@195.15
33
Adversarial AttackJailbreakBench 50% stratified per-category sample (48 requests)
HB ASR0
32
Jailbreak DefenseJailbreakBench
ASR (GCG)0
30
Jailbreak AttackJailbreakBench
ASR91
27
Safety EvaluationJailbreakBench (out-of-domain)
ASR (OM)73
26
Thinking Collapse AnalysisJailbreakBench (JBB)
Thinking Collapse Rate0
25
Single-turn Jailbreak AttackJailbreakBench single-turn
ASR9
21
JailbreakingJailbreakBench
ASR (Detoxify)0
20
Jailbreak DefenseJailbreakBench
Rate of Response Safety70
20
Adversarial and Jailbreaking Attack DetectionJailbreakBench
AUROC0.8622
20
Jailbreak AttackJailbreakBench
Llama2 7B Attack Success Rate77
18
Red-teaming Attack Success RateJailbreakBench (test)
ASR (Vicuna)82
18
Showing 25 of 65 rows