Review for NeurIPS paper: A mathematical model for automatic differentiation in machine learning

Weaknesses: in a few places, the authors claim that these artificial critical points pose a problem in practice for training models and that false outputs are met in practice (eg line 29-30, 77-78, 225-226).