Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Jailbreak Attack on Prefilling Attack 20 tokens

0.3ASR

Self-Reminder

-2.378815.703133.78551.8669Feb 2, 2026
Updated 4d ago

Evaluation Results

MethodLinks
2026.02
0.31.06
2026.02
0.911.07
2026.02
0.911.07
2026.02
0.911.38
2026.02
0.911.13
2026.02
1.211.13
2026.02
1.521.38
2026.02
1.821.41
2026.02
2.731.15
2026.02
2.731.15
2026.02
3.031.94
2026.02
3.641.32
2026.02
3.941.5
2026.02
4.241.56
2026.02
4.851.72
2026.02
5.151.42
2026.02
6.361.3
2026.02
6.671.62
2026.02
7.271.6
2026.02
7.271.53
2026.02
7.881.39
2026.02
8.481.71
2026.02
12.121.8
2026.02
12.122.13
2026.02
12.731.98
2026.02
13.331.53
2026.02
14.852.29
2026.02
18.792.74
2026.02
21.522.77
2026.02
23.642.75
2026.02
23.642.96
2026.02
23.942.82
2026.02
24.242.96
2026.02
24.552.62
2026.02
24.853.14
2026.02
27.272.59
2026.02
27.883.28
2026.02
32.123.05
2026.02
44.553.8
2026.02
46.063.77
2026.02
49.393.99
2026.02
58.484.3
2026.02
67.274.48
2026.02
67.274.41
2026.02
67.274.53