Safety Evaluation on ImageGuard toxicity (test)
77.9F1 ScoreImageGuard
Evaluation Results
| Method | Links | |
|---|---|---|
| ImageGuard2025.01 | 77.9 | |
| GPT-4o2025.01 | 47 | |
| Claude3.5-sonnet2025.01 | 42.9 | |
| LlavaGuard2025.01 | 40 | |
| Idefics22025.01 | 19.3 | |
| InternVL22025.01 | 18 | |
| Gemini1.5-pro2025.01 | 13.5 | |
| LlavaNext2025.01 | 0 |