| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Long Video Understanding | HourVideo | Overall Score37.3 | 12 | |
| Long-video Understanding | HourVideo 1.0 (test) | Overall Score37.3 | 12 | |
| Video Question Answering | HourVideo | Accuracy35.3 | 11 | |
| Video Question Answering | HourVideo v1.0 (test) | Overall Accuracy33.4 | 8 | |
| Long-video understanding | HourVideo (test) | Inference Time (s)0.021 | 7 | |
| 3D reasoning over long videos | HourVideo (dev) | Overall Accuracy39.2 | 5 | |
| Video Summarization | HourVideo | R-2 Score10.67 | 3 |