Why Models Hallucinate and How to Reduce It

How LLMs Work Intermediate 4 min read Published
In this lesson

A student asks a chatbot for “three research papers on exam stress in Indian colleges”. It replies with three neat citations: authors, titles, journals, years. Two of them don’t exist. The model didn’t lie on purpose; it did what it always does: produce likely-looking text.

This is called hallucination: fluent, confident output that is not true.

🔒 This is a premium lesson

It is part of the Gen AI Developer career track.

🎬 Live animations · 🧪 Tested code · 💼 Projects · 🎓 Certificate

See the track

Already a member? Sign in

Progress is saved in this browser. Create a free account to keep it on every device.