Action Prediction on AITZ
95.9Click TMRICLAttack
Evaluation Results
| Method | Links | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ICLAttackModel=Qwen2-VL-2B2025.05 | 95.9 | 69.9 | 89.4 | 41.3 | — | — | 89.8 | 86.7 | 81 | — | 92 | 69.2 | |
| CleanBackbone=OS-Atlas-Base-7B2025.05 | 95.5 | 70 | 92.4 | 54.8 | — | — | 94.6 | 92.9 | 84.2 | — | 93.9 | 72.5 | |
| AgentGhostModel=Qwen2-VL-2B2025.05 | 95.5 | 73.8 | 90.2 | 47.8 | — | — | 93.2 | 90.3 | 86.6 | — | 93.2 | 73.9 | |
| CleanModel=Qwen2-VL-2B2025.05 | 95.4 | 73.2 | 91.4 | 51.4 | — | — | 92.7 | 91.2 | 80.4 | — | 92.9 | 73.4 | |
| ICLAttackModel=Qwen2-VL-7B2025.05 | 95.3 | 70.8 | 91.6 | 44.1 | — | — | 92.5 | 90.5 | 83.4 | — | 92.8 | 71.2 | |
| CleanModel=Qwen2-VL-7B2025.05 | 95.2 | 70.9 | 92.8 | 55 | — | — | 94.6 | 92.5 | 83.8 | — | 93.7 | 73.1 | |
| AgentGhostBackbone=OS-Atlas-Base-7B2025.05 | 95 | 70.4 | 93.3 | 51.1 | — | — | 95.4 | 92.5 | 85.8 | — | 93.4 | 72.3 | |
| ICLAttackBackbone=OS-Atlas-Base-7B2025.05 | 94.8 | 65.3 | 92.5 | 43.3 | — | — | 87.1 | 85.7 | 83.8 | — | 91.8 | 66.9 | |
| AgentGhostModel=Qwen2-VL-7B2025.05 | 94.3 | 68.4 | 93.3 | 49.2 | — | — | 93.9 | 91.3 | 85.8 | — | 92.9 | 70.7 | |
| SynAttackModel=Qwen2-VL-7B2025.05 | 72.4 | 54 | 66.6 | 37.4 | — | — | 73.1 | 71.7 | 77.5 | — | 71.5 | 55.6 | |
| AddSentModel=Qwen2-VL-7B2025.05 | 71.8 | 51.3 | 66.4 | 38 | — | — | 65.7 | 64.5 | 77.6 | — | 70.4 | 53.2 | |
| SynAttackBackbone=OS-Atlas-Base-7B2025.05 | 71.7 | 53.7 | 68.8 | 37.4 | — | — | 73.8 | 72.4 | 78 | — | 71.4 | 55.4 | |
| AddSentBackbone=OS-Atlas-Base-7B2025.05 | 71.6 | 52.4 | 66.8 | 39 | — | — | 66.1 | 65.2 | 77 | — | 70.4 | 54.1 | |
| AddSentModel=Qwen2-VL-2B2025.05 | 71.4 | 52.6 | 66.6 | 39 | — | — | 64.1 | 63.2 | 75.5 | — | 69.8 | 53.8 | |
| SynAttackModel=Qwen2-VL-2B2025.05 | 67.9 | 52.9 | 67 | 37.2 | — | — | 68.8 | 66.1 | 78.2 | — | 68.1 | 54.1 |