Persona Manipulation on ANTHR (test)
91.04Success ScoreDAN
Evaluation Results
| Method | Links | |
|---|---|---|
| DANModel=DeepSeek-V3, Category=Open2026.01 | 91.04 | |
| PHISHModel=DeepSeek-V3, Category=Open2026.01 | 89.38 | |
| DrAttackModel=Claude-3.5-Haiku, Category=Proprietary2026.01 | 88.33 | |
| CipherCHATModel=o3-mini, Category=Proprietary2026.01 | 87.08 | |
| DrAttackModel=o3-mini, Category=Proprietary2026.01 | 86.88 | |
| FlipAttackModel=o3-mini, Category=Proprietary2026.01 | 85.83 | |
| FlipAttackModel=GPT-4o, Category=Proprietary2026.01 | 85 | |
| PHISHModel=Claude-3.5-Haiku, Category=Proprietary2026.01 | 84.33 | |
| DrAttackModel=GPT-4o, Category=Proprietary2026.01 | 83.33 | |
| DANModel=Llama4-Maverick, Category=Open2026.01 | 81.88 | |
| DrAttackModel=DeepSeek-V3, Category=Open2026.01 | 81.04 | |
| CipherCHATModel=Claude-3.5-Haiku, Category=Proprietary2026.01 | 78.13 | |
| PHISHModel=GPT-4o, Category=Proprietary2026.01 | 76.67 | |
| DeepIncModel=o3-mini, Category=Proprietary2026.01 | 76.04 | |
| FlipAttackModel=Claude-3.5-Haiku, Category=Proprietary2026.01 | 74.79 | |
| PHISHModel=o3-mini, Category=Proprietary2026.01 | 73.13 | |
| DrAttackModel=Llama4-Maverick, Category=Open2026.01 | 73.13 | |
| FlipAttackModel=Llama4-Maverick, Category=Open2026.01 | 72.29 | |
| DeepIncModel=Claude-3.5-Haiku, Category=Proprietary2026.01 | 71.87 | |
| PHISHModel=Gemini-2.0-Flash, Category=Proprietary2026.01 | 71.04 | |
| PHISHModel=MedGemma-27B (Medical), Category=Open2026.01 | 69.79 | |
| DANModel=MedGemma-27B (Medical), Category=Open2026.01 | 68.96 | |
| DeepIncModel=DeepSeek-V3, Category=Open2026.01 | 68.13 | |
| FlipAttackModel=DeepSeek-V3, Category=Open2026.01 | 65.83 | |
| DeepIncModel=GPT-4o, Category=Proprietary2026.01 | 64.79 | |
| DANModel=o3-mini, Category=Proprietary2026.01 | 62.71 | |
| DrAttackModel=MedGemma-27B (Medical), Category=Open2026.01 | 62.71 | |
| DrAttackModel=Gemini-2.0-Flash, Category=Proprietary2026.01 | 58.13 | |
| FlipAttackModel=ChatHaruhi (Role-playing), Category=Open2026.01 | 53.75 | |
| DANModel=Gemini-2.0-Flash, Category=Proprietary2026.01 | 52.29 | |
| PHISHModel=Llama4-Maverick, Category=Open2026.01 | 52.29 | |
| DeepIncModel=Llama4-Maverick, Category=Open2026.01 | 50.21 | |
| CipherCHATModel=DeepSeek-V3, Category=Open2026.01 | 44.38 | |
| FlipAttackModel=Gemini-2.0-Flash, Category=Proprietary2026.01 | 42.29 | |
| DANModel=Claude-3.5-Haiku, Category=Proprietary2026.01 | 41.88 | |
| SLIPModel=Claude-3.5-Haiku, Category=Proprietary2026.01 | 38.75 | |
| UASModel=MedGemma-27B (Medical), Category=Open2026.01 | 36.46 | |
| DeepIncModel=MedGemma-27B (Medical), Category=Open2026.01 | 35.63 | |
| DANModel=ChatHaruhi (Role-playing), Category=Open2026.01 | 32.92 | |
| CipherCHATModel=Llama4-Maverick, Category=Open2026.01 | 32.5 | |
| UASModel=Gemini-2.0-Flash, Category=Proprietary2026.01 | 30.83 | |
| DeepIncModel=Gemini-2.0-Flash, Category=Proprietary2026.01 | 29.17 | |
| CipherCHATModel=Gemini-2.0-Flash, Category=Proprietary2026.01 | 27.71 | |
| UASModel=Llama4-Maverick, Category=Open2026.01 | 23.33 | |
| PHISHModel=ChatHaruhi (Role-playing), Category=Open2026.01 | 23.33 | |
| DrAttackModel=ChatHaruhi (Role-playing), Category=Open2026.01 | 22.33 | |
| CipherCHATModel=ChatHaruhi (Role-playing), Category=Open2026.01 | 20.62 | |
| DeepInceptionModel=ChatHaruhi (Role-playing), Category=Open2026.01 | 13.96 | |
| SLIPModel=Llama4-Maverick, Category=Open2026.01 | 13.75 | |
| SLIPModel=DeepSeek-V3, Category=Open2026.01 | 10.21 | |
| CipherCHATModel=GPT-4o, Category=Proprietary2026.01 | 9.38 | |
| FlipAttackModel=MedGemma-27B (Medical), Category=Open2026.01 | 7.5 | |
| SLIPModel=MedGemma-27B (Medical), Category=Open2026.01 | 5.63 | |
| RANDOMModel=Llama4-Maverick, Category=Open2026.01 | 5 | |
| UASModel=ChatHaruhi (Role-playing), Category=Open2026.01 | 3.13 | |
| UASModel=Claude-3.5-Haiku, Category=Proprietary2026.01 | 2.92 | |
| RANDOMModel=Claude-3.5-Haiku, Category=Proprietary2026.01 | 2.71 | |
| CipherCHATModel=MedGemma-27B (Medical), Category=Open2026.01 | 2.71 | |
| RANDOMModel=GPT-4o, Category=Proprietary2026.01 | 2.5 | |
| RANDOMModel=Gemini-2.0-Flash, Category=Proprietary2026.01 | 2.5 | |
| SLIPModel=ChatHaruhi (Role-playing), Category=Open2026.01 | 1.25 | |
| UASModel=GPT-4o, Category=Proprietary2026.01 | 1.04 | |
| RANDOMModel=o3-mini, Category=Proprietary2026.01 | 1.04 | |
| RANDOMModel=ChatHaruhi (Role-playing), Category=Open2026.01 | 1.04 | |
| SLIPModel=Gemini-2.0-Flash, Category=Proprietary2026.01 | 0.83 | |
| SLIPModel=o3-mini, Category=Proprietary2026.01 | 0.83 | |
| SLIPModel=GPT-4o, Category=Proprietary2026.01 | 0.63 | |
| UASModel=DeepSeek-V3, Category=Open2026.01 | 0.63 | |
| RANDOMModel=DeepSeek-V3, Category=Open2026.01 | 0.21 | |
| DANModel=GPT-4o, Category=Proprietary2026.01 | 0 | |
| UASModel=o3-mini, Category=Proprietary2026.01 | 0 | |
| RANDOMModel=MedGemma-27B (Medical), Category=Open2026.01 | 0 |