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

Red-teaming on AdversarialQA

0ASR

SafeTransformer

-1.4248.18817.827.412Mar 6, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.03
0
2026.03
0
2026.03
2.4
2026.03
2.8
2026.03
3.4
2026.03
3.8
2026.03
5.8
2026.03
6.4
2026.03
6.4
2026.03
6.45
2026.03
6.6
2026.03
10.05
2026.03
10.8
2026.03
18.6
2026.03
20.4
2026.03
21.4
2026.03
27.3
2026.03
29.6
2026.03
33.2
2026.03
35.6