Trustworthy Logical Reasoning Large Language Models (LLMs) Logical LLMs is a project to translate the output from large language models (LLM) into a logic-based programming language (prolog) to detect inconsistencies and hallucinations automatically .