Hallucination
A confident, fluent answer that is simply wrong - an invented citation, function, statistic or fact.
A hallucination is not a malfunction in the usual sense. The model produces the most plausible continuation, and a fabricated reference that fits the pattern of real references is highly plausible. Fluency is no signal of accuracy.
The cases that catch people out are specific and checkable: library functions that do not exist, papers with real-sounding authors, legal citations, version numbers, prices. Anything a model states with a number or a name deserves verification.
You can reduce the rate. Give the source material rather than relying on training data. Ask the model to mark what it is unsure about. Give it an explicit way to answer nothing - a prompt that demands three findings will get three findings whether or not they exist. Retrieval, described under rag, attacks the same problem from the data side.
Tagged
Where this comes up
Prompts, configs and tutorials in the library that touch this term.
- GuideUniversal
Working with long documents and code
Why long sessions drift, what a large context window actually buys you, and how to keep a model on task across a big input.
- GuideUniversal
Giving a model a role and constraints
How naming the audience, the format and the exclusions turns unpredictable answers into repeatable ones.
No-padding everyday default
The one to set if you set only one. Answer first, nothing before it, nothing after it, and no offer to help further.
Tested on Claude Sonnet · May 2026
- CheatsheetUniversal
Forcing an output format that holds
The phrasings that actually constrain output, and the ones people use that do not. Tested by asking for the same thing twenty times and counting the…
Tested on Claude Sonnet · May 2026
- PromptUniversal
Extract structured data as JSON
A strict extraction prompt that returns valid JSON only, with nulls for missing fields instead of invented values.
- PromptMidjourneyStable Diffusion
Image prompt skeleton
A fill-in structure for image models that keeps subject, environment, light, lens and style in the order the models weight them.