Retrieval Augmented Generation (RAG)

RAG is a grounding technique popularized by Meta AI researchers to improve the quality and relevance of AI-generated content by allowing LLMs to access external knowledge sources other than the model's training data.