| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| VSR | UniMRG | Overall Accuracy73.9 | 17 | 4mo ago | |
| CV-Bench | TIGeR | 2D Relational Score93.85 | 9 | 4mo ago | |
| MultihopSpatial | Thinking with Grounding Box | Accuracy37.68 | 8 | 1mo ago | |
| EmbSpatial | Gemma3-27B-IT | Accuracy62.09 | 8 | 1mo ago | |
| VSR zeroshot | Gemma3-27B-IT | Accuracy69.25 | 8 | 1mo ago |