Spatial Reasoning on OrthoMind-3D (OOD)
38.7Block Count Accuracy
3ViewSense-4B-rl-slack
Updated 2mo ago
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
2026.03 | 38.7 | 70.2 | 50.9 | 76.1 | |
2026.03 | 33.2 | 71.1 | 49.1 | 74.3 | |
2026.03 | 33.1 | 65.9 | 54.6 | 34.8 | |
2026.03 | 31.1 | 62.1 | 32.4 | 72.5 | |
2026.03 | 26.8 | 58.7 | 47.2 | 20.1 | |
2026.03 | 25.5 | 64.7 | 56.5 | 55 | |
2026.03 | 23.4 | 61.7 | 46.2 | 19.2 | |
2026.03 | 21.2 | 62.5 | 51.8 | 16.5 | |
2026.03 | 21.2 | 57.8 | 50.9 | 46.7 | |
2026.03 | 19.6 | 43 | 22.2 | 15.6 |