Loading TutorKit...
Hallucinations
A clear-eyed, practical guide to AI hallucinations: what they are, why a next-token predictor produces false or unsupported claims, why "just say I don't know" is harder than it sounds, when they're most likely, how to detect one, and how to reduce and work around them.
Questions in this set
- What is an AI hallucination?
- Why do models hallucinate?
- If it doesn't know, why doesn't it just say "I don't know"?
- What do hallucinations look like?
- When are hallucinations most likely?
- How can you tell whether an answer is hallucinated?
- How do you reduce hallucinations?
- Are bigger and newer models fixing this?
- How should you work with a model that hallucinates?
TutorWant me to explain it differently?
AI concepts can be dense. Tell me what's confusing and I'll find a new analogy.