We generally don't care enough to test it since it would be a significate waste of resources that's why we have models in the trillions of parameters instead of

We generally don't care enough to test it since it would be a significate waste of resources that's why we have models in the trillions of parameters instead of continuing to train smaller models smaller models hit a wall and generally stop improving and start overfitting and generalization starts to degrade it's less of a theorem, but an observation that has occured during training and documented here and there across many papers.