On the Compositional Generalization of Multimodal LLMs for Medical Imaging

Open in new window