Spatial Reasoning on RoboSpatial (Conf, Comp, Cont Metrics)
82.11ConfidenceTIGeR
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| TIGeR2025.10 | 82.11 | 82.86 | 32.79 | |
| GLM-4.1V-Thinking2025.10 | 81.3 | 43.8 | 8.24 | |
| Gemini-2.5-Pro2025.10 | 77.24 | 68.57 | 40.2 | |
| GPT-4o2025.10 | 77.2 | 61.9 | 25.3 | |
| SpatialBot-3BParameters=3B2025.10 | 72.36 | 22.86 | 4.88 | |
| RoboPoint-13BParameters=13B2025.10 | 69.9 | 21.9 | 41.3 | |
| SpaceLLaVA-13BParameters=13B2025.10 | 61 | 38.1 | 16 | |
| NVILA-8BParameters=8B2025.10 | 59.35 | 23.81 | 5.51 | |
| Qwen-2.5-VL-7BParameters=7B2025.10 | 49.59 | 66.67 | 13.61 |