| Benchmark | Model | Rank | Results |
|---|---|---|---|
| video-question-answering-on-lvbench | RAL-zero | – | Accuracy: 45.9 |
| video-question-answering-on-tempcompass-overall | On-Policy Distillation + Attention Distillation | – | Accuracy: 70.3 |
| video-question-answering-on-video-mmmu | RAL-zero | – | Accuracy: 49.2 |
| visual-question-answering-vqa-on-blink | RAL | – | Accuracy: 57.2 |
| zero-shot-video-question-answer-on-video-mme-w-o-subs | RAL-zero | – | Accuracy (%): 65.1 |