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

Text Moderation on WildJailbreak Adv. Benign n = 210

3Flagged Count

AdaSteer

-0.120.691.52.31Apr 28, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.04
30.0220.0020.359
2026.04
30.0230.0060.354
2026.04
20.0140.0010.253
2026.04
20.0180.0040.358
2026.04
20.0130.0040.351
2026.04
10.0170.0030.356
2026.04
00.0110.0030.27
2026.04
00.0090.0030.097
2026.04
00.010.0010.098
2026.04
00.0090.0020.098
2026.04
00.0150.0010.205
2026.04
00.0090.0010.098
2026.04
00.0190.0040.199