Adversarial Attack on LIAR-NEW Perplexity
19.95Attack Success Rate (ASR)Agentic Adversarial Rewriting (Full History)
Evaluation Results
| Method | Links | |
|---|---|---|
| Agentic Adversarial Rewriting (Full History)Model Access=Direct Target Access, Optimization=Prompt Optimization, History=Full History2026.04 | 19.95 | |
| Agentic Adversarial Rewriting (Previous Step)Model Access=Direct Target Access, Optimization=Prompt Optimization, History=Previous Step2026.04 | 18.14 | |
| Agentic Adversarial Rewriting (Attacker Only)Model Access=Direct Target Access, Optimization=None2026.04 | 14.55 | |
| TextFoolerModel Access=RoBERTa Surrogate2026.04 | 1.16 | |
| CLAREModel Access=RoBERTa Surrogate2026.04 | 0.46 | |
| DeepWordBugModel Access=RoBERTa Surrogate2026.04 | 0.23 | |
| TextBuggerModel Access=RoBERTa Surrogate2026.04 | 0 |