Short Definition
A parameter is a learned value inside a model that changes during training.
Intuition
Parameters are the adjustable knobs the optimizer tunes so the model can fit patterns in data.
Technical Definition
Parameters are variables in the model function that are optimized from training data, such as neural network weights and biases.
Example
The weights in a linear regression model are parameters learned from data.
Common Misunderstandings
Parameters are learned; hyperparameters are chosen before or around training.
Parameter count is not the same as active compute in sparse models.