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

L3J

Benchmarks

Task NameDataset NameSOTA ResultTrend
Jailbreak Defense EvaluationL3J
Strong-Reject Score (SR)3.341
21
Jailbreak DetectionL3J
Accuracy98.3
13
Showing 2 of 2 rows