Hallucination Rate Measurement on Medical Diagnostic Hallucination Dataset All
82.3Auto Detection RateQwen2-VL
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Qwen2-VL2026.05 | 82.3 | 80.2 | |
| InternVL22026.05 | 76.6 | 71.4 | |
| Gemini 3 Flash2026.05 | 71.2 | 43 | |
| GPT-5.12026.05 | 71.1 | 46.8 | |
| LLaVA-NeXT2026.05 | 67.1 | 74.7 | |
| BiomedCLIP2026.05 | 61.9 | 36.4 |