Generative Multimodal Models are In-Context Learners | IEEE Conference Publication | IEEE Xplore