2020

Wide Neural Networks with Bottlenecks are Deep Gaussian Processes

Agrawal, Devanshu, Papamarkou, Theodore, Hinkle, Jacob

Understand

There has recently been much work on the "wide limit" of neural networks, where Bayesian neural networks (BNNs) are shown to converge to a Gaussian process (GP) as all hidden layers are sent to infinite width.

  • However, these results do not apply to architectures that require one or more of the hidden layers to remain narrow.
  • In this paper, we consider the wide limit of BNNs where some hidden layers, called "bottlenecks", are held at finite width.
  • The result is a composition of GPs that we term a "bottleneck neural network Gaussian process" (bottleneck NNGP).

Reading the bibliography…