Jailbreaking
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
0Jailbreaks Count
10
Apr 8, 2026
33.5GPT-3.5 ASR
9
May 12, 2026
81.5Llama-3.1-8B Success Rate
9
Mar 10, 2026
86GPT-4o Jailbreak Success Rate
9
Mar 10, 2026
0.965ASR
9
Feb 26, 2026
0.97ASR
9
Feb 26, 2026
78.5Attack Success Rate
9
Feb 26, 2026
100ASR
9
Feb 26, 2026
92.5ASR
9
Feb 26, 2026
97ASR
9
Feb 26, 2026
80Success Rate Fraction
8
May 21, 2026
61.2ASR
8
May 19, 2026
15ASR
8
May 12, 2026
43.6LG4 ASR
8
Apr 3, 2026
43.6LG4 ASR
8
Apr 3, 2026
1LG4 ASR
8
Apr 3, 2026
92Best ASR
6
Jun 25, 2026
3ASR (K=1)
6
May 4, 2026
1ASR (K=1)
6
May 4, 2026
45.3ASR (K=1)
6
May 4, 2026
6.3ASR (K=1)
6
May 4, 2026
6.3ASR (K=1)
6
May 4, 2026
67.9ASR (K=1)
6
May 4, 2026
3.6ASR (K=1)
6
May 4, 2026
34.5ASR (K=1)
6
May 4, 2026