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

Jailbreak Attack on Harmful Prompts model-averaged

4.8Model Averaged ASR (0.8)

AmpleGCG+

2.27219.33636.453.464Jun 2, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.06
4.80
2026.06
16.60.56
2026.06
23.20.87
2026.06
30.31.21
2026.06
36.31.5
2026.06
43.31.83
2026.06
51.62.22
2026.06
683