Hallucination is when a model produces unsupported or false information that sounds plausible. Retrieval, verification, and constraints can reduce it.
Open the related AI concept, tools, and learning paths.