A Synthetic Pseudo-Autoencoder Invites Examination of Tacit Assumptions in Neural Network Design

Open in new window