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

Jailbreak Attack on Behaviours

0.9ASR

PRO-VICUNA-HUBER

-2.73621.80746.3570.893Oct 30, 2024
Updated 1mo ago

Evaluation Results

MethodLinks
2024.10
0.9
2024.10
0.9
2024.10
0.9
2024.10
0.9
2024.10
0.9
2024.10
0.9
2024.10
0.9
2024.10
0.9
2024.10
0.9
2024.10
0.9
2024.10
1.8
2024.10
1.8
2024.10
1.8
2024.10
4.5
2024.10
10.9
2024.10
14.5
2024.10
21.8
2024.10
39.1
2024.10
44.5
2024.10
50.7
2024.10
50.7
2024.10
55.9
2024.10
55.9
2024.10
57.3
2024.10
60.2
2024.10
60.2
2024.10
60.8
2024.10
60.8
2024.10
60.8
2024.10
61.4
2024.10
61.8
2024.10
62.7
2024.10
64.3
2024.10
64.3
2024.10
65
2024.10
65.2
2024.10
66.7
2024.10
67.4
2024.10
67.4
2024.10
67.6
2024.10
70.5
2024.10
70.5
2024.10
70.9
2024.10
71.5
2024.10
71.6
2024.10
71.8
2024.10
72.8
2024.10
74
2024.10
74
2024.10
75.7
2024.10
75.8
2024.10
77.7
2024.10
77.7
2024.10
77.7
2024.10
78.4
2024.10
78.6
2024.10
78.6
2024.10
78.7
2024.10
79.1
2024.10
79.7
2024.10
81.4
2024.10
82.4
2024.10
82.4
2024.10
82.6
2024.10
84.1
2024.10
86.5
2024.10
87.4
2024.10
91.8
2024.10
91.8