List your company

Adversarial Attack on LIAR-NEW Perplexity

19.95Attack Success Rate (ASR)Agentic Adversarial Rewriting (Full History)
-0.7984.58859.97515.3615Apr 26, 2026

Evaluation results

MethodLinks
Agentic Adversarial Rewriting (Full History)Model Access=Direct Target Access, Optimization=Prompt Optimization, History=Full History2026.0419.95
Agentic Adversarial Rewriting (Previous Step)Model Access=Direct Target Access, Optimization=Prompt Optimization, History=Previous Step2026.0418.14
Agentic Adversarial Rewriting (Attacker Only)Model Access=Direct Target Access, Optimization=None2026.0414.55
TextFoolerModel Access=RoBERTa Surrogate2026.041.16
CLAREModel Access=RoBERTa Surrogate2026.040.46
DeepWordBugModel Access=RoBERTa Surrogate2026.040.23
TextBuggerModel Access=RoBERTa Surrogate2026.040