SpatiaLab: Can Vision-Language Models Perform Spatial Reasoning in the Wild? Paper • 2602.03916 • Published 3 days ago • 10