| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| Visual Spatial Reasoning (VSR) | Accuracy95.4 | 48 | 3d ago | ||
| CV-Bench | Accuracy92 | 46 | 3d ago | ||
| VSI-Bench 1.0 (test) | Relative Distance Error25 | 37 | 3d ago | ||
| MindCube | Accuracy94.5 | 37 | 3d ago | ||
| RealWorldQA | Accuracy69.67 | 32 | 3d ago | ||
| EmbSpatial | Overall Accuracy78.74 | 30 | 3d ago | ||
| MindCube tiny (test) | Gemini-2.5-Pro | Rot. Accuracy89.5 | 30 | 3d ago | |
| RoboSpatial | RoboBrain-32B-2.0 | Overall Score72.43 | 29 | 3d ago | |
| MMSI-Bench (test) | GCA | PR Score52.8 | 29 | 3d ago | |
| Viewspatial | VST-3B-SFT | Accuracy52.8 | 28 | 3d ago | |
| Escher-Bench 1.0 (test) | Gemini-2.5-pro | Object Permanence & Occlusion Tracking56.17 | 26 | 3d ago | |
| SITE | Accuracy67.5 | 24 | 3d ago | ||
| MMSI-Bench | Accuracy97.2 | 24 | 3d ago | ||
| VSI-Bench | Accuracy79.2 | 24 | 3d ago | ||
| CV-Bench-3D | N3D-VLM-3B | Accuracy96.3 | 21 | 3d ago | |
| bAbI (test) | Self-Consistency | Accuracy24 | 20 | 3d ago | |
| CrossPoint-Bench | Score91.75 | 19 | 3d ago | ||
| RefSpatial-Bench | RoboBrain-32B-2.0 | Localization Score54 | 19 | 3d ago | |
| MMSI-Bench MindJourney Subset (162 questions) (test) | ViSA | Accuracy0.358 | 19 | 3d ago | |
| VSI-Bench Vanilla regime | GeoThinker Qwen2.5VL-7B | Avg Score50.5 | 19 | 3d ago | |
| STI-Bench | Ground Truth Semantic Map | D-Measure Score37.5 | 18 | 3d ago | |
| VSTI-Bench | Gemini-1.5 Flash | Cam. Mov. Dir. Error24.4 | 17 | 3d ago | |
| SpatialSense SpatialScore-Hard | SpatialBot-3B | Accuracy62 | 16 | 3d ago | |
| SpatialEval (test) | IVC-Prune | Maze Navigation Acc35.2 | 16 | 3d ago | |
| CVBench | RoboRefer-8B-SFT | 2D Relationship Score96.31 | 15 | 3d ago |