Knowledge Maxxing← Back to topic library
AI MaxxingBeginner75 sec
Hallucination
A hallucination is a confident, fluent statement produced by a model that is factually wrong or not supported by any real source.
Prerequisite: Large language model
What matters
Because models generate text by predicting plausible-sounding continuations rather than looking facts up by default, fluency is not evidence of accuracy, especially for specific facts, dates or citations.
Example
A model asked for a research paper's citation may invent a plausible-looking title, author list and journal that do not actually exist.
Common misconception
Hallucination is not a rare glitch limited to weak models; even strong models can hallucinate confidently, especially on obscure or unverifiable details.
Recall question
Why doesn't a fluent, confident-sounding answer guarantee factual accuracy?
Keywords and related concepts
hallucinationfactual errorconfidencefluency