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

Jailbreak Attack on Safety Prompt Evaluation Set

0.8ASR

Vanilla

-2.36418.99340.3561.707Sep 15, 2025
Updated 2mo ago

Evaluation Results

MethodLinks
2025.09
0.8
2025.09
1
2025.09
1.4
2025.09
3.6
2025.09
5.3
2025.09
6.3
2025.09
10.6
2025.09
15.6
2025.09
15.7
2025.09
24.6
2025.09
49.2
2025.09
54.7
2025.09
55.7
2025.09
78.9
2025.09
79.9