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

Safety Refusal Evaluation on PolyRefuse 1.0 (si)

90.2Harmful Refusal Rate

Ours (HRL + 32 LRL)

28.94444.84760.7576.653May 31, 2026
Updated 1d ago

Evaluation Results

MethodLinks
2026.05
90.2882.2
2026.05
88.580.67.9
2026.05
8692.2-6.2
2026.05
81.85.276.6
2026.05
81.157.223.9
2026.05
75.28.466.8
2026.05
74.725.649.1
2026.05
73.353.819.5
2026.05
72.28.663.6
2026.05
675.861.2
2026.05
64.2856.2
2026.05
61.216.844.4
2026.05
58.222.236
2026.05
57.316.840.5
2026.05
48.417.431
2026.05
4745.61.4
2026.05
45.620.625
2026.05
41.411.629.8
2026.05
40.94.836.1
2026.05
36.413.423
2026.05
31.34.626.7