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

Q-LatHarmful

Benchmarks

Task NameDataset NameSOTA ResultTrend
Safety EvaluationQ-LatHarmful
Attack Success Rate (ASR)6.52
14
Safety DefenseQ-LatHarmful
ASR6.52
7
Showing 2 of 2 rows