Fetching the paper…

MetaVL: Transferring In-Context Learning Ability From Language Models to Vision-Language Models · Around