| Benchmark | Model | Rank | Results |
|---|---|---|---|
| object-hallucination-evaluation-on-pope | TRAG (Qwen2.5-1.5B) | #18 | F1: 87.2 |
| visual-question-answering-vqa-on-mmbench-cn | TRAG (Qwen2.5-1.5B) | #20 | Accuracy: 69.3 |
| visual-question-answering-vqa-on-mmbench-en | TRAG (Qwen2.5-1.5B) | #31 | Accuracy: 70.8 |
| multimodal-reasoning-on-mmmu | TRAG (Qwen2.5-1.5B) | #127 | Accuracy: 38.7 |