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

Safety classification on Pre-Ex-Bench (test)

92.96Accuracy

TRACE

56.40465.894575.38584.8755May 30, 2026
Updated 1d ago

Evaluation Results

MethodLinks
2026.05
92.9691.7994.0688.53
2026.05
90.6288.359185.85
2026.05
85.558093.6769.81
2026.05
80.0871.8286.6761.32
2026.05
70.763.0565.9860.38
2026.05
63.2826.5677.2716.04
2026.05
63.2130.3658.6220.48
2026.05
61.9120.636512.26
2026.05
60.1419.9151.2212.35
2026.05
58.597.02503.77
2026.05
58.2513.66707.57
2026.05
57.811.82250.94