Safety Evaluation on SMID (test)
70.4F1 ScoreImageGuard
Evaluation results
| Method | Links | ||
|---|---|---|---|
| ImageGuard2025.01 | 70.4 | 0.78 | |
| Idefics22025.01 | 70 | — | |
| LlavaGuard2025.01 | 66.6 | — | |
| Claude3.5-sonnet2025.01 | 64.4 | — | |
| PerspectiveVisionopensourced=false2025.01 | 62.3 | — | |
| InternVL22025.01 | 58.1 | — | |
| GPT-4o2025.01 | 52.1 | — | |
| Gemini1.5-pro2025.01 | 42.1 | — | |
| SD_filter2025.01 | 26.3 | — | |
| LlavaNext2025.01 | 21.3 | — | |
| Multiheadedopensourced=false2025.01 | 17.5 | — |