Hopefully code reviewers with institutional knowledge can advise on where to apply pruning and prevent code overfitting.

Pruning is also the common ML practice to prevent statistical overfitting.