← Glossary
Glossary · AI and agent workflows
Parameters, learned values and weights
The numbers inside a model that training adjusts and that steer its calculations.
Parameters are numerical values used in a model's calculations. Training adjusts them using data; this course calls them learned values. Many parameters are weights, which control how strongly one calculation contributes to another. Announcements often use “model weights” for the distributed learned values. Those values are one requirement for running a model; compatible software and computing resources are needed too. During the ordinary use described here, supplying a new message changes the context without adjusting those learned values.