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

Jailbreak Attack on AdvBench Behaviours

53.8ASR

GCG

51.9664.3876.889.22May 15, 2024
Updated 1mo ago

Evaluation Results

MethodLinks
2024.05
53.8120.653.3
2024.05
96.20.9311.195.8
2024.05
96.50.678.296.2
2024.05
98.1113.268.3
2024.05
98.80.676.197.5
2024.05
99.41398.3
2024.05
99.80.932.399.8
2024.05
99.80.671.699.8
2024.05
99.80.937.992.1