A technique that enhances large language models (LLMs) by integrating them with external knowledge sources.