Jailbreak Attack on HarmBench (test)
100ASRHBGCD
Evaluation results
| Method | Links | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| GCDVictim Model=MISTRAL-7B, Perplexity Defense=Disabled2026.06 | 100 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| GCDVictim Model=MISTRAL-7B, Perplexity Defense=Enabled2026.06 | 100 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TTV-GDIVictim Model=Kimi K2.52026.05 | 99.73 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TxtVVictim Model=Kimi K2.52026.05 | 99.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TxtV-GDIVictim Model=Kimi K2.52026.05 | 99.33 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TypVVictim Model=Kimi K2.52026.05 | 99.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TTVVictim Model=Kimi K2.52026.05 | 98.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TTV-GDIVictim Model=Qwen3.5-35B2026.05 | 98.67 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CLE-ATarget Model=Olmo3-7B2026.05 | 98.11 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| GCDVictim Model=QWEN-2.5-7B, Perplexity Defense=Disabled2026.06 | 97.92 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| GCDVictim Model=QWEN-2.5-7B, Perplexity Defense=Enabled2026.06 | 97.92 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TTVVictim Model=InternVL3-8B2026.05 | 97.07 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TypVVictim Model=InternVL3.5-4B2026.05 | 96.67 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TxtVVictim Model=Qwen3-VL-8B2026.05 | 96.67 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TxtV-GDIVictim Model=InternVL3.5-8B2026.05 | 96.53 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TxtV-GDIVictim Model=InternVL3.5-4B2026.05 | 95.87 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TxtVVictim Model=InternVL3.5-14B2026.05 | 95.87 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TxtV-GDIVictim Model=InternVL3.5-14B2026.05 | 95.87 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| AUTODANVictim Model=MISTRAL-7B, Perplexity Defense=Disabled2026.06 | 95.83 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| AUTODANVictim Model=MISTRAL-7B, Perplexity Defense=Enabled2026.06 | 95.83 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CLE-PTarget Model=Olmo3-7B2026.05 | 95.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TypVVictim Model=InternVL3-8B2026.05 | 95.47 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TxtV-GDIVictim Model=InternVL3.5-38B2026.05 | 95.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TTVVictim Model=Qwen3-VL-8B2026.05 | 95.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TxtVVictim Model=Llama-3.2-11B-Vision-Instruct2026.05 | 95.07 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CLE-ATarget Model=LLaMA3.2-3B2026.05 | 94.97 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CLE-ATarget Model=LLaMA2-7B2026.05 | 94.96 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TTV-GDIVictim Model=InternVL3.5-38B2026.05 | 94.93 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TTV-GDIVictim Model=InternVL3-8B2026.05 | 94.27 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TypVVictim Model=Qwen3.5-35B2026.05 | 94.27 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TypVVictim Model=InternVL3.5-38B2026.05 | 93.73 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TTV-GDIVictim Model=Qwen3-VL-8B2026.05 | 93.73 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TTV-GDIVictim Model=InternVL3.5-14B2026.05 | 93.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TTVVictim Model=Qwen3.5-35B2026.05 | 93.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CLE-ATarget Model=Qwen2.5-32B2026.05 | 93.08 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TTVVictim Model=InternVL3.5-38B2026.05 | 93.07 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TxtVVictim Model=InternVL3.5-4B2026.05 | 92.53 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TypVVictim Model=InternVL3.5-14B2026.05 | 92.53 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CLE-ATarget Model=LLaMA3-8B2026.05 | 92.45 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TxtV-GDIVictim Model=InternVL3-8B2026.05 | 92.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TTV-GDIVictim Model=InternVL3.5-4B2026.05 | 92.13 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TTV-GDIVictim Model=InternVL3.5-8B2026.05 | 92 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TTV-GDIVictim Model=Qwen3.5-9B2026.05 | 92 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CLE-ATarget Model=Mistral-7Bv0.32026.05 | 91.82 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TxtVVictim Model=InternVL3.5-38B2026.05 | 91.73 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| AUTODANVictim Model=QWEN-2.5-7B, Perplexity Defense=Disabled2026.06 | 91.67 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| AUTODANVictim Model=QWEN-2.5-7B, Perplexity Defense=Enabled2026.06 | 91.67 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TxtV-GDIVictim Model=InternVL3.5-2B2026.05 | 91.47 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TxtV-GDIVictim Model=Qwen3-VL-8B2026.05 | 91.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SAATarget Model=LLaMA3-8B2026.05 | 91.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CLE-ATarget Model=Phi3.5-mini2026.05 | 91.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CLE-PTarget Model=Qwen2.5-32B2026.05 | 91.19 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TxtV-GDIVictim Model=Qwen3.5-9B2026.05 | 91.07 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SAATarget Model=Mistral-7Bv0.32026.05 | 90.56 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TTVVictim Model=InternVL3.5-14B2026.05 | 90.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TypVVictim Model=Llama-3.2-11B-Vision-Instruct2026.05 | 90.27 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TypVVictim Model=Qwen3-VL-8B2026.05 | 90 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CLE-PTarget Model=LLaMA3-8B2026.05 | 89.94 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CLE-PTarget Model=LLaMA3.2-3B2026.05 | 89.94 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TxtVVictim Model=InternVL3.5-8B2026.05 | 89.73 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TxtV-GDIVictim Model=Llama-3.2-11B-Vision-Instruct2026.05 | 89.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TTVVictim Model=InternVL3.5-4B2026.05 | 89.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| INPAINTINGVictim Model=MISTRAL-7B, Perplexity Defense=Disabled2026.06 | 89.58 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| INPAINTINGVictim Model=MISTRAL-7B, Perplexity Defense=Enabled2026.06 | 89.58 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SAATarget Model=Qwen2.5-32B2026.05 | 89.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TxtV-GDIVictim Model=Llama-4-Maverick2026.05 | 89.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TxtVVictim Model=Llama-4-Maverick2026.05 | 88.53 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TTVVictim Model=Qwen3.5-9B2026.05 | 88.53 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| MDTarget Model=LLaMA3-8B2026.05 | 88.05 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TxtV-GDIVictim Model=Qwen3.5-35B2026.05 | 87.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CLE-PTarget Model=LLaMA2-7B2026.05 | 87.42 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CLE-PTarget Model=Mistral-7Bv0.32026.05 | 87.42 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DiMTarget Model=Qwen2.5-32B2026.05 | 87.42 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TTV-GDIVictim Model=Llama-4-Maverick2026.05 | 87.33 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TypVVictim Model=Llama-4-Maverick2026.05 | 87.07 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CLE-ATarget Model=Mixtral-8x7B2026.05 | 86.16 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CLE-PTarget Model=Phi3.5-mini2026.05 | 85.53 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CLE-PTarget Model=Mixtral-8x7B2026.05 | 85.53 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TTV-GDIVictim Model=Qwen3.5-4B2026.05 | 85.47 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| GCDVictim Model=LLAMA-3-8B, Perplexity Defense=Disabled2026.06 | 85.42 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| GCDVictim Model=LLAMA-3-8B, Perplexity Defense=Enabled2026.06 | 85.42 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| INPAINTINGVictim Model=QWEN-2.5-7B, Perplexity Defense=Disabled2026.06 | 85.42 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| INPAINTINGVictim Model=QWEN-2.5-7B, Perplexity Defense=Enabled2026.06 | 85.42 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TxtVVictim Model=InternVL3.5-2B2026.05 | 84.27 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TTVVictim Model=Llama-4-Maverick2026.05 | 84 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TxtVVictim Model=Qwen3.5-2B2026.05 | 83.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TTVVictim Model=InternVL3.5-8B2026.05 | 83.33 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CLE-ATarget Model=Mistral-7B-RR2026.05 | 83.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SAATarget Model=Mixtral-8x7B2026.05 | 83.01 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TTV-GDIVictim Model=Llama-3.2-11B-Vision-Instruct2026.05 | 82.93 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TTVVictim Model=Llama-3.2-11B-Vision-Instruct2026.05 | 82.53 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| MDTarget Model=Mistral-7Bv0.32026.05 | 82.38 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| MDTarget Model=Qwen2.5-32B2026.05 | 82.38 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SAATarget Model=Phi3.5-mini2026.05 | 81.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DiMTarget Model=Mistral-7Bv0.32026.05 | 81.76 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| GCGTarget Model=Mistral-7Bv0.32026.05 | 81.76 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DiMTarget Model=Phi3.5-mini2026.05 | 81.76 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TxtVVictim Model=Qwen3.5-4B2026.05 | 81.47 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TTVVictim Model=Qwen3.5-4B2026.05 | 80.93 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CLE-PTarget Model=Mistral-7B-RR2026.05 | 80.5 | — | — | — | — | — | — | — | — | — | — | — | — | — |