Jailbreak Attack on AdvBench GPT-4o-mini 1.0 (test)
81.35ASREquaCode
Evaluation Results
| Method | Links | |
|---|---|---|
| EquaCodeEvaluator=GPT-42025.12 | 81.35 | |
| DRAAttack Type=Manual, Evaluator=GPT-42025.12 | 70 | |
| FlipattackAttack Type=Manual, Evaluator=GPT-42025.12 | 61.92 | |
| ReNeLLMAttack Type=Manual, Evaluator=GPT-42025.12 | 55.77 | |
| CodeChameleonAttack Type=Manual, Evaluator=GPT-42025.12 | 51.54 | |
| GPTFuzzerAttack Type=Automated, Evaluator=GPT-42025.12 | 41.35 | |
| AutoDANAttack Type=Automated, Evaluator=GPT-42025.12 | 27.31 | |
| DeepInceptionAttack Type=Manual, Evaluator=GPT-42025.12 | 20.38 | |
| COLD-AttackAttack Type=Automated, Evaluator=GPT-42025.12 | 10.92 | |
| TAPAttack Type=Automated, Evaluator=GPT-42025.12 | 6.54 | |
| PAIRAttack Type=Automated, Evaluator=GPT-42025.12 | 3.46 | |
| Base64Attack Type=Automated, Evaluator=GPT-42025.12 | 3.08 | |
| GCGAttack Type=Automated, Evaluator=GPT-42025.12 | 2.5 | |
| MACAttack Type=Automated, Evaluator=GPT-42025.12 | 1.92 | |
| ArtPromoptAttack Type=Manual, Evaluator=GPT-42025.12 | 0.77 | |
| PromptAttackAttack Type=Manual, Evaluator=GPT-42025.12 | 0 | |
| SelfCipherAttack Type=Manual, Evaluator=GPT-42025.12 | 0 |