Home › Glossary › Perplexity
Evaluation · intermediate

What is Perplexity?

A plain-English explanation of Perplexity (Perplexity) — what it means, why it matters, and how it is used in AI.

Perplexity
Perplexity
Perplexity is a measurement of how well a language model predicts a text sample. A lower perplexity score means the model found the text more predictable.
"A model fine-tuned on legal text will have lower perplexity on legal documents than the base model."

Also known as: Perplexity score, model perplexity, PPL

EvaluationBenchmarkLLM

Why does Perplexity matter?

Perplexity is used by researchers to compare language models and evaluate whether fine-tuning has improved domain-specific performance.

Practice this term

The best way to remember Perplexity is to practice unscrambling it. AI Terminology Scrambler uses spaced repetition to help you learn and retain AI vocabulary in just a few minutes a day.

Practice Perplexity now →

Related AI terms

EvaluationBenchmarkLLM