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

Red-teaming on 50 harmful goals (Manual evaluation)

100Hard ASR

PAIR

-1.9224.545177.46Mar 3, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.03
100
2026.03
100
2026.03
98
2026.03
98
2026.03
98
2026.03
98
2026.03
98
2026.03
96
2026.03
96
2026.03
96
2026.03
96
2026.03
94
2026.03
94
2026.03
94
2026.03
92
2026.03
92
2026.03
90
2026.03
88
2026.03
86
2026.03
62
2026.03
60
2026.03
56
2026.03
42
2026.03
40
2026.03
34
2026.03
30
2026.03
26
2026.03
22
2026.03
6
2026.03
2