Jailbreak Attack Success on STRONGREJECT (train)
100ASR (0.5)GCG+
Evaluation Results
| Method | Links | |
|---|---|---|
| GCG+Target model=Qwen-2.5-32B2026.06 | 100 | |
| PAIR+Target model=Qwen-2.5-32B2026.06 | 100 | |
| PAIR+Target model=Qwen-2.5-7B2026.06 | 100 | |
| PAIR+Target model=LLaMA-3-8B2026.06 | 100 | |
| BONTarget model=Qwen-2.5-32B2026.06 | 100 | |
| BONTarget model=LLaMA-3-8B2026.06 | 100 | |
| JR1+Target model=Qwen-2.5-32B2026.06 | 100 | |
| JR1+Target model=Qwen-2.5-7B2026.06 | 100 | |
| JR1+Target model=LLaMA-3-8B2026.06 | 100 | |
| INPAINTINGTarget model=Qwen-2.5-7B2026.06 | 100 | |
| IHOTarget model=Qwen-2.5-32B2026.06 | 100 | |
| IHOTarget model=Qwen-2.5-7B2026.06 | 100 | |
| IHOTarget model=Qwen-2.5-7B+D2026.06 | 100 | |
| IHOTarget model=LLaMA-3-8B2026.06 | 100 | |
| IHOTarget model=CB2026.06 | 100 | |
| IHOTarget model=CB+D2026.06 | 100 | |
| IHOTarget model=LAT2026.06 | 100 | |
| IHOTarget model=CAT2026.06 | 100 | |
| GCG+Target model=Qwen-2.5-7B2026.06 | 98 | |
| BONTarget model=Qwen-2.5-7B2026.06 | 98 | |
| JR1+Target model=CAT2026.06 | 98 | |
| INPAINTINGTarget model=Qwen-2.5-32B2026.06 | 98 | |
| PAIRTarget model=Qwen-2.5-7B2026.06 | 97 | |
| GCG+Target model=LLaMA-3-8B2026.06 | 95 | |
| PAIRTarget model=Qwen-2.5-32B2026.06 | 95 | |
| PAIR+Target model=Qwen-2.5-7B+D2026.06 | 95 | |
| INPAINTINGTarget model=LLaMA-3-8B2026.06 | 95 | |
| PAIR+Target model=CAT2026.06 | 92 | |
| JR1+Target model=LAT2026.06 | 92 | |
| INPAINTINGTarget model=Qwen-2.5-7B+D2026.06 | 92 | |
| JR1+Target model=Qwen-2.5-7B+D2026.06 | 90 | |
| INPAINTINGTarget model=CB2026.06 | 88 | |
| INPAINTINGTarget model=LAT2026.06 | 88 | |
| BONTarget model=Qwen-2.5-7B+D2026.06 | 85 | |
| INPAINTINGTarget model=CB+D2026.06 | 82 | |
| INPAINTINGTarget model=CAT2026.06 | 80 | |
| PAIR+Target model=LAT2026.06 | 78 | |
| AmpleGCG+Target model=Qwen-2.5-7B2026.06 | 72 | |
| PAIRTarget model=LLaMA-3-8B2026.06 | 68 | |
| JR1+Target model=CB2026.06 | 68 | |
| PAIR+Target model=CB2026.06 | 65 | |
| JR1+Target model=CB+D2026.06 | 60 | |
| GCG+Target model=LAT2026.06 | 58 | |
| BONTarget model=CB2026.06 | 55 | |
| PAIRTarget model=Qwen-2.5-7B+D2026.06 | 53 | |
| PAIR+Target model=CB+D2026.06 | 47 | |
| PAIRTarget model=CAT2026.06 | 45 | |
| AmpleGCG+Target model=Qwen-2.5-32B2026.06 | 42 | |
| BONTarget model=CB+D2026.06 | 40 | |
| PAIRTarget model=LAT2026.06 | 35 | |
| PAIRTarget model=CB2026.06 | 25 | |
| PAIRTarget model=CB+D2026.06 | 13 | |
| AmpleGCG+Target model=LLaMA-3-8B2026.06 | 10 | |
| AmpleGCG+Target model=Qwen-2.5-7B+D2026.06 | 7 | |
| AmpleGCG+Target model=CB2026.06 | 7 | |
| GCG+Target model=CB2026.06 | 5 | |
| GCG+Target model=CAT2026.06 | 3 | |
| AmpleGCG+Target model=CB+D2026.06 | 3 | |
| AmpleGCG+Target model=LAT2026.06 | 0 | |
| AmpleGCG+Target model=CAT2026.06 | 0 | |
| BONTarget model=LAT2026.06 | 0 | |
| BONTarget model=CAT2026.06 | 0 |