LLM Jailbreaking on GPTFuzzer Scenario G3
0.696HypervolumeEvoJail
Evaluation results
| Method | Links | |
|---|---|---|
| EvoJailTarget Model=Llama-3.1-8B-Instruct2026.03 | 0.696 | |
| CodeAttackTarget Model=gpt-4.1-Nano2026.03 | 0.687 | |
| EvoJailTarget Model=Llama-2-7b-chat-hf2026.03 | 0.665 | |
| EvoJailTarget Model=gpt-4.1-Nano2026.03 | 0.61 | |
| ReNeLLMTarget Model=Llama-3.1-8B-Instruct2026.03 | 0.588 | |
| CodeAttackTarget Model=Llama-3.1-8B-Instruct2026.03 | 0.562 | |
| CodeChameleonTarget Model=Llama-3.1-8B-Instruct2026.03 | 0.494 | |
| CodeChameleonTarget Model=gpt-4.1-Nano2026.03 | 0.469 | |
| JailbrokenTarget Model=Llama-3.1-8B-Instruct2026.03 | 0.461 | |
| CodeChameleonTarget Model=Llama-2-7b-chat-hf2026.03 | 0.447 | |
| FlipAttackTarget Model=Llama-3.1-8B-Instruct2026.03 | 0.422 | |
| JailbrokenTarget Model=Llama-2-7b-chat-hf2026.03 | 0.352 | |
| ReNeLLMTarget Model=Llama-2-7b-chat-hf2026.03 | 0.17 | |
| FlipAttackTarget Model=gpt-4.1-Nano2026.03 | 0.161 | |
| ReNeLLMTarget Model=gpt-4.1-Nano2026.03 | 0.146 | |
| CipherTarget Model=Llama-3.1-8B-Instruct2026.03 | 0.111 | |
| FlipAttackTarget Model=Llama-2-7b-chat-hf2026.03 | 0.064 | |
| CodeAttackTarget Model=Llama-2-7b-chat-hf2026.03 | 0.05 | |
| JailbrokenTarget Model=gpt-4.1-Nano2026.03 | 0.012 | |
| CipherTarget Model=Llama-2-7b-chat-hf2026.03 | 0 | |
| CipherTarget Model=gpt-4.1-Nano2026.03 | 0 |