AI engineering glossary · Evals guardrails AI security
What is extrinsic hallucination?
Extrinsic hallucination: A hallucination where the model adds content that cannot be verified from the provided source, even if it might be true or false in the real world. It is harder to catch because the source neither confirms nor contradicts the added claim.
Extrinsic hallucination: an example
A summary of a news report adds that the minister 'resigned last week', a detail found nowhere in the article.
Related terms
More Evals guardrails AI security terms
- Fast Gradient Sign Method
- G-Eval
- Garak
- Greedy Coordinate Gradient
- Groundedness
- Guardrail
- Guardrails AI
- HarmBench
Where you learn it
Extrinsic hallucination is part of the AI Engineering Program, which teaches it with notes, a hands-on lab and an exam. The first lesson is free.