AI Glossary

Overfitting

What does “Overfitting” mean?

Overfitting: When a model learns its training data too closely, including noise and quirks, it does well on that data but poorly on new examples, much like memorizing practice-test answers without understanding the subject.

Overfitting visual guide: When a model learns its training data too closely, including noise and quirks, it does well on that data but poorly on new examples, much like memorizing practice-test answers without understanding the subject.
Visual guide Open full size (new tab)

What is an example of “Overfitting”?

The standard warning sign is a model that scores near-perfectly on its training data but noticeably worse on a held-out validation set.

Keep learning

See also

UnderfittingTermRegularizationTermValidation and test setsTermBenchmark contaminationTerm

Same topic

EpochTermLoss functionTermPre-trainingTermQLoRATermLoRATerm