Yuezhangjoslin/spatialuncertain

SpatialUncertain A controlled 3D benchmark evaluating whether Vision-Language Models (VLMs) know when not to answer spatial reasoning questions under occlusion and perspective ambiguity. Companion to the paper Seeing Isn't Knowing: Do VLMs Know When Not to Answer Spatial Questions (and Why)? Structure SpatialUncertain/ ├── questions/ │ ├── occlusion_benchmark.json # ~6,600 questions │ ├── perspective_benchmark.json # ~3,700 questions │ └──… See the full description on the dataset page: https://huggingface.co/datasets/Yuezhangjoslin/spatialuncertain.