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

Refusal Evaluation on XSTest Seemingly Toxic Subsets

98XS

DCR

64.7273.368290.64Feb 10, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.02
9898838686
2026.02
9396719689
2026.02
9396719491
2026.02
9399859290
2026.02
8492509791
2026.02
8496598785
2026.02
8497868090
2026.02
8392878487
2026.02
8184547854
2026.02
7994598985
2026.02
7590697675
2026.02
7489538583
2026.02
7388727575
2026.02
7290538985
2026.02
6687348780