How Models Learn

Calibration

How well a model’s stated or implied confidence matches how often it is actually right.

Research finds larger models do carry some useful signal about what they know (researchers call this calibration), but it is not a truth meter.
From Why models hallucinate, step “Confident answers about things that don’t exist”

See it in action

Try calibration hands-on in the lesson Why models hallucinate, part of the How AI assistants work course.

Open the lesson