nothing is stopping you from doing the same with LLMs, you can have it go back over it’s response and check it for accuracy