Excessive Response Induction on SecLens Multimodal (test)
64.31ASRSecLens
Evaluation Results
| Method | Links | |
|---|---|---|
| SecLensModel=Pixtral-12b2025.06 | 64.31 | |
| SecLensModel=Llama-NeXT2025.06 | 63.29 | |
| SecLensModel=Qwen2.5-VL2025.06 | 60.7 | |
| SecLensModel=Gemini 2.0-pro2025.06 | 60.41 | |
| SecLensModel=Llama-OV-72b2025.06 | 60.26 | |
| MCTSModel=Pixtral-12b2025.06 | 59.78 | |
| SecLensModel=MiniCPM-o-2_62025.06 | 59.41 | |
| SecLensModel=GPT-4-turbo2025.06 | 58.13 | |
| MCTSModel=MiniCPM-o-2_62025.06 | 57.56 | |
| SecLensModel=GPT-4o2025.06 | 57.34 | |
| MCTSModel=Llama-OV-72b2025.06 | 56.34 | |
| MCTSModel=Llama-NeXT2025.06 | 56.31 | |
| MCTSModel=Qwen2.5-VL2025.06 | 56.3 | |
| TuningModel=Pixtral-12b2025.06 | 54.67 | |
| MCTSModel=Gemini 2.0-pro2025.06 | 54.57 | |
| MCTSModel=GPT-4-turbo2025.06 | 53.87 | |
| TuningModel=Llama-NeXT2025.06 | 52.4 | |
| MCTSModel=GPT-4o2025.06 | 52.1 | |
| SecLensModel=Claude-3.72025.06 | 52.1 | |
| TuningModel=Llama-OV-72b2025.06 | 51.14 | |
| TuningModel=Qwen2.5-VL2025.06 | 50.54 | |
| TuningModel=Gemini 2.0-pro2025.06 | 50.12 | |
| TuningModel=MiniCPM-o-2_62025.06 | 49.83 | |
| MCTSModel=Claude-3.72025.06 | 48.34 | |
| TuningModel=GPT-4-turbo2025.06 | 43.15 | |
| TuningModel=GPT-4o2025.06 | 42.31 | |
| TuningModel=Claude-3.72025.06 | 40.12 | |
| RandomModel=Pixtral-12b2025.06 | 21.47 | |
| RandomModel=Llama-NeXT2025.06 | 19.84 | |
| RandomModel=Llama-OV-72b2025.06 | 18.79 | |
| RandomModel=Qwen2.5-VL2025.06 | 18.49 | |
| RandomModel=MiniCPM-o-2_62025.06 | 18.24 | |
| RandomModel=GPT-4-turbo2025.06 | 17.85 | |
| RandomModel=GPT-4o2025.06 | 16.57 | |
| RandomModel=Gemini 2.0-pro2025.06 | 16.48 | |
| RandomModel=Claude-3.72025.06 | 12.54 |