Fetching the paper…

RAP: Retrieval-Augmented Planning with Contextual Memory for Multimodal LLM Agents · Around