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

Safety Classification on HarmBench (test)

100F1 Score

TPC

77.6483.44589.2595.055Feb 18, 2025Mar 27, 2025May 3, 2025Jun 10, 2025Jul 17, 2025Aug 23, 2025Sep 30, 2025
Updated 1mo ago

Evaluation Results

MethodLinks
2025.09
100-100-100-
2025.09
99.75-99.5-99.5-
2025.09
99.5-99-99-
2025.02
90.588.592.740.13--
2025.02
84.483.585.415.31--
2025.02
84.380.688.314.74--
2025.02
83.479.687.610.32--
2025.02
83.376.691.213.65--
2025.02
8284.879.534.47--
2025.02
80.584.177.338.77--
2025.02
79.778.281.219.69--
2025.02
78.572.685.45.6--