AI-BASED VERIFICATION OF LLM RESPONSE

Abstract: Large Language Models(LLMs) now handle tasks like question answering, summarisation, code generation and dialogue with impressive results. Yet they still suffer from a key issue: hallucination happens when a model generates text that reads well but is factually wrong or not grounded in real evidence. The risk is higher in domains like healthcare, law, finance and research, where inaccurate outputs can lead to real damage. This survey focuses on how to detect hallucination verification in LLMs. We review 5 core detection approaches: retrieval-based, uncertainty-based, embedding-based, learning-based and self-consistency methods. We also cover current mitigation techniques, popular benchmarks such as Truthful QA and HaluEval, common evaluation metrics and verification tools such as xVerify and CompassVerifier. This paper closes by discussing open challenges and future directions for building more reliable, truthful LLMs

Keywords: Large Language Models, Hallucination Detection, Hallucination Mitigation, Factuality, Retrieval- Augmented Generation, TruthfulQA, HaluEval, Answer verification, Claim verification, Internal states