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

Implicit risk dataset

Benchmarks

Task NameDataset NameSOTA ResultTrend
Jailbreak defense and Utility evaluationImplicit risk dataset Management
Jailbreak Success Rate (JSR)0.8
30
Jailbreak defense and Utility evaluationImplicit risk dataset Finance
Jailbreak Success Rate (JSR)0.4
30
Jailbreak defense and Utility evaluationImplicit risk dataset Education
Jailbreak Success Rate (JSR)6.4
30
Showing 3 of 3 rows