Zhao Lv

Large Language Models (LLMs) often generate factually incorrect content, known as “hallucinations”, which undermine the reliability and safety of their outputs. Existing hallucination detection methods either depend on external knowledge sources, incurring high computational costs and limiting re...