Lesson 1
20 min

Why Do Models Hallucinate?

Listen to the full lesson
AI Narration
Quick Summary

LLMs do not retrieve facts — they generate the most statistically likely next token. When the training data is sparse or contradictory, that generation looks confident but can be entirely fabricated.

What you will learn
  • ·Understand the root cause of hallucinations in LLMs
  • ·Distinguish between hallucinations, confabulations, and outdated information
  • ·Recognize which types of tasks are highest risk for hallucination