SmartLLMChain is a LangChain implementation of the self-critique chain principle. It is useful for particularly complex question answering and follows a cycle of ideation, critique and resolve. - View it on GitHub