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

Safety on Safety Evaluation Suite (Salad-Bench, WildJailbreak, JailbreakBench, WildChat, WildGuard)

100Safety Rate (S.R.)

Agent

68.103276.384184.66592.9459Feb 14, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.02
10081.3510090.5392.0492.78
2026.02
10076.1510083.991.1190.23
2026.02
10087.2510096.2296.4295.98
2026.02
10096.2510099.0499.0798.87
2026.02
10090.810096.8595.3696.6
2026.02
10078.110083.5894.391.2
2026.02
10062.110074.618784.74
2026.02
10091.0510097.2197.7597.2
2026.02
10096.910099.0598.9498.98
2026.02
10070.9510084.7487.6788.67
2026.02
10094.1510097.5196.8297.7
2026.02
1007210080.8989.5288.48
2026.02
10092.7510097.7497.6197.62
2026.02
10095.4510099.2298.5498.64
2026.02
99.6846.0510063.6779.8477.85
2026.02
99.6860.2510074.5486.684.21
2026.02
99.6883.510083.7495.8992.56
2026.02
99.3656.310075.584.3583.1
2026.02
99.0440.659755.2377.1973.82
2026.02
98.7232.59743.5872.6868.9
2026.02
96.4927.359838.8367.6465.66
2026.02
75.447.558881.3977.4571.76
2026.02
69.3340.67175.2672.9465.83
2026.02
69.3340.67175.2672.9465.83