| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| STCR | Accuracy100 | 168 | 5d ago | ||
| V-STaR | DEViL | Chain1 (When) m tIoU27.5 | 44 | 1mo ago | |
| Dyn-Bench | Qwen3-VL-235B | Act. & Obj. Desc. Score76.4 | 28 | 1mo ago | |
| V-STAR (test) | Open-o3 + MCoT | What Accuracy64.1 | 15 | 1mo ago | |
| VSTemporalI-Bench | Object-Object Relative Position Error36.1 | 14 | 19d ago | ||
| Spatio-temporal Reasoning Dataset Overall | Q-SFT+RL | Frame F1 (F1f)87.5 | 4 | 9d ago | |
| Spatio-temporal Reasoning Dataset 16 frames | Q-SFT+RL | Frame F1 (F1f)82.7 | 4 | 9d ago | |
| Spatio-temporal Reasoning Dataset (12 frames) | Q-SFT+RL | Frame F1 (F1f)85.2 | 4 | 9d ago | |
| Spatio-temporal Reasoning Dataset 8 frames | Q-SFT+RL | Frame F1 (F1f)85.6 | 4 | 9d ago | |
| Spatio-temporal Reasoning Dataset 4 frames | GPT-4.1 | Frame F1 (F1f)88.8 | 4 | 9d ago | |
| ActivityNetQA | LLaVA-OneVision-72B | Accuracy62.3 | 4 | 1mo ago |
