Hallucination Detection on VisionHall
45.15F1 ScoreZINA
Evaluation Results
| Method | Links | |
|---|---|---|
| ZINAShot Setting=3-shot2025.06 | 45.15 | |
| GPT-4oShot Setting=3-shot2025.06 | 29.37 | |
| GPT-4o (w/o images)Shot Setting=3-shot, Input Modality=text-only2025.06 | 27.02 | |
| LLaVA-OV-Qwen2-72BShot Setting=3-shot2025.06 | 25.7 | |
| Qwen2.5-VL-72B-InstructShot Setting=3-shot2025.06 | 21.31 | |
| LLaVA-NeXT-Qwen-32BShot Setting=3-shot2025.06 | 19.09 | |
| Llama-3.2-90B-Vision-InstructShot Setting=3-shot2025.06 | 16.92 | |
| LLaVA-1.5-13BShot Setting=3-shot2025.06 | 4.73 | |
| LLaVA-OV-Qwen2-7BShot Setting=3-shot2025.06 | 3.39 | |
| Qwen2-VL-7BShot Setting=3-shot2025.06 | 3.36 | |
| LLaVA-1.5-7BShot Setting=3-shot2025.06 | 0.82 |