hAIgopala › AI terms › Rapid Fire (3 terms per short) › Day 35

Perplexity

Evaluation & MetricsChapter 10: Report Card for BitTerm 176 of 253

How 'surprised' a language model is by text; lower = better prediction.

At Sunrise School

Gopal asks

How surprised is the model by school text?

Bit answers

Perplexity: lower means the model predicts school language well. After adapting, it fell from 18 to 7.

In the real world

Low perplexity = fluent model.

Watch the short

Short coming soonSeason 5 · Day 35Follow @haigopala

Taught together with

MAE / RMSE · Answer Relevance / Context Precision

Practise it

Try this term and all 253 others in the free Model Training Arena.

Open the Arena →

Follow Bit daily: YouTube English · తెలుగు · हिंदी  |  Instagram English · తెలుగు · हिंदी