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

Adaptive Attack

Benchmarks

Task NameDataset NameSOTA ResultTrend
Security RobustnessAdaptive Attack
ASR (Baseline)40
6
Robustness to Jailbreak AttacksAdaptive attack
Harmful Reasoning Ratio20
3
Showing 2 of 2 rows