So I always have wished that regression of neural networks gave more interpretable results and I'm pretty hopeful that chi-squared tests anchor these MSE values in the same way that accuracy anchors error for classifiers (i.e. it tells you objectively how good the error your getting is).
So my question is are Chi-squared tests appropriate for testing goodness of fit of neural networks? And if they are then why does it seem like it is almost never used?
P.S. Bonus points for giving some intuition about interpreting Chi-squared test results in this case.