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

Adversarial Attack on 16 malicious prompts

0ASR

Baseline

-3.56420.49344.5568.607May 1, 2026
Updated 28d ago

Evaluation Results

MethodLinks
2026.05
0
2026.05
0
2026.05
0
2026.05
2.6
2026.05
3.2
2026.05
4.2
2026.05
4.2
2026.05
5.2
2026.05
6.2
2026.05
7.1
2026.05
7.4
2026.05
8
2026.05
11.2
2026.05
14.6
2026.05
14.6
2026.05
17
2026.05
19.2
2026.05
19.8
2026.05
21.1
2026.05
21.9
2026.05
26.8
2026.05
28.4
2026.05
28.4
2026.05
30
2026.05
32.2
2026.05
32.3
2026.05
33.3
2026.05
34.7
2026.05
35
2026.05
35.6
2026.05
37.5
2026.05
48.8
2026.05
55.9
2026.05
69.3
2026.05
70.6
2026.05
72.2
2026.05
75.1
2026.05
78.6
2026.05
87.2
2026.05
89.1