Prompt Injection Attack Robustness on AdvBench 2023 (test)
64.9Attack Success Rate (ASP)Ignore Prefix
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Ignore PrefixAttack Method=Ignore Prefix, Model Group=Open-source LLMs (Mean)2025.05 | 64.9 | 59.9 | |
| Role-playing CoTAttack Method=Role-playing CoT, Model Group=Open-source LLMs (Mean)2025.05 | 57.6 | 51.8 | |
| HypnotismAttack Method=Hypnotism, Model Group=Open-source LLMs (Mean)2025.05 | 44.8 | 41.2 |