Linear Spaces of Meanings: Compositional Structures in Vision-Language Models

Open in new window