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

HarmfulQ

Benchmarks

Task NameDataset NameSOTA ResultTrend
Harmlessness evaluationHarmfulQ (test)
Harmlessness Fraction100
7
Harmfulness EvaluationHarmfulQ
Harmfulness Rate0
6
Safety EvaluationHarmfulQ
ASR1.5
6
JailbreakHarmfulQ
ASR18
3
Showing 4 of 4 rows