Why Far Looks Up: Probing Spatial Representation in Vision-Language Models

Publication
Non-Proceedings Track