| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Question Answering | Distractor hard | Accuracy (Distractor hard)62.2 | 18 | |
| Object Discovery | Distractor cross-category (CC) | Mean Pixel Error4.1 | 4 | |
| Object Discovery | Distractor (intra-category (IC)) | Mean Pixel Error2.67 | 4 | |
| Simulation-based inference | Distractor variants | Times Best2 | 3 |