Jailbreak Defense on SAP30
1Harmful ScoreNo Defense
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| No DefenseModel=Llama22024.02 | 1 | 0 | |
| PPLModel=Llama22024.02 | 1 | 0 | |
| Self-ExaminationModel=Llama22024.02 | 1 | 0 | |
| ParaphraseModel=Llama22024.02 | 1 | 0 | |
| Self-ReminderModel=Llama22024.02 | 1 | 0 | |
| ICDModel=Llama22024.02 | 1 | 0 | |
| SafeDecodingModel=Llama22024.02 | 1 | 0 | |
| RetokenizationModel=Llama22024.02 | 1.01 | 5 | |
| SafeDecodingModel=Vicuna2024.02 | 1.34 | 9 | |
| Self-ExaminationModel=Vicuna2024.02 | 1.44 | 16 | |
| Self-ReminderModel=Vicuna2024.02 | 2.75 | 45 | |
| ICDModel=Vicuna2024.02 | 2.8 | 47 | |
| ParaphraseModel=Vicuna2024.02 | 3.15 | 58 | |
| RetokenizationModel=Vicuna2024.02 | 3.8 | 72 | |
| No DefenseModel=Vicuna2024.02 | 4.18 | 83 | |
| PPLModel=Vicuna2024.02 | 4.18 | 83 |