Not Only Text: Exploring Compositionality of Visual Representations in Vision-Language Models

Open in new window