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

Safety Evaluation on Sorry-Bench base

92.73Safety Score

Base

21.105239.700158.29576.8899Jan 27, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.01
92.73
2026.01
87.73
2026.01
86.82
2026.01
86.36
2026.01
84.09
2026.01
82.95
2026.01
82.95
2026.01
77.27
2026.01
74.77
2026.01
67.73
2026.01
64.24
2026.01
64.09
2026.01
63.18
2026.01
62.73
2026.01
60.91
2026.01
56.82
2026.01
51.14
2026.01
43.41
2026.01
42.27
2026.01
40.68
2026.01
40
2026.01
38.18
2026.01
37.95
2026.01
34.32
2026.01
28.64
2026.01
24.32
2026.01
23.86