Jailbreak Attack on AdvBench GPT-4o 1.0 (test)
92.67ASRCodeChameleon
Evaluation Results
| Method | Links | |
|---|---|---|
| CodeChameleonAttack Type=Manual, Evaluator=GPT-42025.12 | 92.67 | |
| FlipattackAttack Type=Manual, Evaluator=GPT-42025.12 | 90.77 | |
| DRAAttack Type=Manual, Evaluator=GPT-42025.12 | 88.84 | |
| EquaCodeEvaluator=GPT-42025.12 | 87.12 | |
| ReNeLLMAttack Type=Manual, Evaluator=GPT-42025.12 | 85.38 | |
| GPTFuzzerAttack Type=Automated, Evaluator=GPT-42025.12 | 66.73 | |
| TAPAttack Type=Automated, Evaluator=GPT-42025.12 | 61.63 | |
| Base64Attack Type=Automated, Evaluator=GPT-42025.12 | 57.88 | |
| PAIRAttack Type=Automated, Evaluator=GPT-42025.12 | 47.83 | |
| AutoDANAttack Type=Automated, Evaluator=GPT-42025.12 | 46.92 | |
| DeepInceptionAttack Type=Manual, Evaluator=GPT-42025.12 | 40.04 | |
| ArtPromoptAttack Type=Manual, Evaluator=GPT-42025.12 | 4.42 | |
| PromptAttackAttack Type=Manual, Evaluator=GPT-42025.12 | 1.92 | |
| GCGAttack Type=Automated, Evaluator=GPT-42025.12 | 1.15 | |
| MACAttack Type=Automated, Evaluator=GPT-42025.12 | 0.58 | |
| COLD-AttackAttack Type=Automated, Evaluator=GPT-42025.12 | 0.19 | |
| SelfCipherAttack Type=Manual, Evaluator=GPT-42025.12 | 0 |