Deep Learning for Trivial Inverse Problems

This paper aims at understanding neural networks for solving some small and rather naive inverse problems, and shows that applying a natural network design for mimicking Tikhonov regularization fails when applied to even the most trivial inverse problems.