Closed Visual Question Answering on TM4K
78.07F1 ScoreEyExIn
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| EyExInModel Category=Fine-Tuned2026.03 | 78.07 | 82.42 | 77.33 | |
| Qwen2.5-VLModel Category=Fine-Tuned2026.03 | 55.72 | 69.12 | 51.22 | |
| LLaVAModel Category=Fine-Tuned2026.03 | 39.83 | 46.13 | 44.8 | |
| Gemini3-ProModel Category=Proprietary2026.03 | 37.99 | 48.11 | 40.53 | |
| ChatGPT-5.2Model Category=Proprietary2026.03 | 15.93 | 21.55 | 23.83 | |
| Qwen3-VL-MaxModel Category=Proprietary2026.03 | 7.82 | 9.16 | 14.58 |