True, but to be a bit pedantic, I wouldn’t call that generalization error (which you don’t have access to while training). Training minimizes some loss function

True, but to be a bit pedantic, I wouldn’t call that generalization error (which you don’t have access to while training). Training minimizes some loss function which is a proxy for generalization error.