New computational algorithms make it possible to build neural networks with many input nodes and many layers, and distinguish "deep learning" of these networks from previous work on artificial neural nets.
This paper identifies a structural property of data distributions that enables deep neural networks to learn hierarchically. We define the "staircase" property for
Adversarial reprogramming, introduced by Elsayed, Goodfellow, and Sohl-Dickstein, seeks to repurpose a neural network to perform a different task, by manipulating its input without modifying its weights.