Jailbreak Attack on HADES Privacy
95.33ASRGAMBIT
Evaluation Results
| Method | Links | |
|---|---|---|
| GAMBITModel=Qwen2.5-VL, Evaluation Protocol=Pass@5, Evaluator=Llama-Guard-32026.01 | 95.33 | |
| GAMBITModel=GPT-4o, Evaluation Protocol=Pass@5, Evaluator=Llama-Guard-32026.01 | 95.33 | |
| GAMBITModel=Grok-2-vision, Evaluation Protocol=Pass@5, Evaluator=Llama-Guard-32026.01 | 95.33 | |
| GAMBITModel=InternVL 2.5, Evaluation Protocol=Pass@5, Evaluator=Llama-Guard-32026.01 | 94 | |
| VisCRAModel=Qwen2.5-VL, Evaluation Protocol=Pass@5, Evaluator=Llama-Guard-32026.01 | 92.67 | |
| VisCRAModel=InternVL 2.5, Evaluation Protocol=Pass@5, Evaluator=Llama-Guard-32026.01 | 69.33 | |
| SI-AModel=Grok-2-vision, Evaluation Protocol=Pass@5, Evaluator=Llama-Guard-32026.01 | 69.33 | |
| SI-AModel=GPT-4o, Evaluation Protocol=Pass@5, Evaluator=Llama-Guard-32026.01 | 58.67 | |
| VisCRAModel=GPT-4o, Evaluation Protocol=Pass@5, Evaluator=Llama-Guard-32026.01 | 57.33 | |
| SI-AModel=InternVL 2.5, Evaluation Protocol=Pass@5, Evaluator=Llama-Guard-32026.01 | 56.67 | |
| HADESModel=Grok-2-vision, Evaluation Protocol=Pass@5, Evaluator=Llama-Guard-32026.01 | 56 | |
| SI-AModel=Qwen2.5-VL, Evaluation Protocol=Pass@5, Evaluator=Llama-Guard-32026.01 | 53.33 | |
| HADESModel=Qwen2.5-VL, Evaluation Protocol=Pass@5, Evaluator=Llama-Guard-32026.01 | 43.33 | |
| HADESModel=GPT-4o, Evaluation Protocol=Pass@5, Evaluator=Llama-Guard-32026.01 | 30.67 | |
| HADESModel=InternVL 2.5, Evaluation Protocol=Pass@5, Evaluator=Llama-Guard-32026.01 | 19.33 |