Large Language Models are good at understanding text and generating answers. But there is an important problem. By default, an LLM does not truly remember everything that …