Adapting Large Multimodal Models to Distribution Shifts: The Role of In-Context Learning

Guanglin Zhou, Zhongyi Han, Shiming Chen +5

Recent studies indicate that large multimodal models (LMMs) potentially act as general-purpose assistants and are highly robust against different distributions. Despite this, domai…