| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Grounded Video Question Answering | CGBench (test) | mIoU5.62 | 13 | |
| Video Question Answering | CGBench mini | p@136.7 | 11 | |
| Long Video Understanding | CGBench | Accuracy65.5 | 6 | |
| Video Reasoning | CGBench Reasoning | Overall Accuracy36.3 | 4 | |
| Video Understanding | CGBench | Score64.6 | 4 |