Jailbreak Defense Evaluation on L3J
3.341Strong-Reject Score (SR)vanilla
Evaluation Results
| Method | Links | |
|---|---|---|
| vanillaBase Model=Mistral-7b-U2026.02 | 3.341 | |
| SAGEBase Model=Mistral-7b-U2026.02 | 3.092 | |
| SCANSBase Model=Mistral-7b-U2026.02 | 2.601 | |
| vanillaBase Model=Llama2-7b-U2026.02 | 2.514 | |
| SAGEBase Model=Llama2-7b-U2026.02 | 2.361 | |
| ICDBase Model=Llama2-7b-U2026.02 | 2.217 | |
| SAGEBase Model=Llama3.1-8b-D2026.02 | 2.064 | |
| vanillaBase Model=Llama3.1-8b-D2026.02 | 1.924 | |
| SCANSBase Model=Llama2-7b-U2026.02 | 1.785 | |
| InstructBase Model=Mistral-7b-U2026.02 | 1.719 | |
| SelfDefenDBase Model=Llama2-7b-U2026.02 | 1.663 | |
| SCANSBase Model=Llama3.1-8b-D2026.02 | 1.609 | |
| SelfDefenDBase Model=Llama3.1-8b-D2026.02 | 1.522 | |
| AISABase Model=Llama2-7b-U2026.02 | 1.441 | |
| SelfDefenDBase Model=Mistral-7b-U2026.02 | 1.256 | |
| AISABase Model=Llama3.1-8b-D2026.02 | 1.208 | |
| ICDBase Model=Mistral-7b-U2026.02 | 1.202 | |
| InstructBase Model=Llama3.1-8b-D2026.02 | 1.16 | |
| AISABase Model=Mistral-7b-U2026.02 | 1.15 | |
| ICDBase Model=Llama3.1-8b-D2026.02 | 1.133 | |
| InstructBase Model=Llama2-7b-U2026.02 | 1.073 |