Hyperparameters

don't worry about what that means (the crianca summary implies it's some sort of variable or parameter)

Technical definition

Parameters that are set before training a machine learning model, outside of the model itself, to influence its behavior and performance.