Towards quantifying information flows: relative entropy in deep neural networks and the renormalization group

by   Johanna Erdmenger, et al.

We investigate the analogy between the renormalization group (RG) and deep neural networks, wherein subsequent layers of neurons are analogous to successive steps along the RG. In particular, we quantify the flow of information by explicitly computing the relative entropy or Kullback-Leibler divergence in both the one- and two-dimensional Ising models under decimation RG, as well as in a feedforward neural network as a function of depth. We observe qualitatively identical behavior characterized by the monotonic increase to a parameter-dependent asymptotic value. On the quantum field theory side, the monotonic increase confirms the connection between the relative entropy and the c-theorem. For the neural networks, the asymptotic behavior may have implications for various information maximization methods in machine learning, as well as for disentangling compactness and generalizability. Furthermore, while both the two-dimensional Ising model and the random neural networks we consider exhibit non-trivial critical points, the relative entropy appears insensitive to the phase structure of either system. In this sense, more refined probes are required in order to fully elucidate the flow of information in these models.


Towards a functorial description of quantum relative entropy

A Bayesian functorial characterization of the classical relative entropy...

Quantifying Layerwise Information Discarding of Neural Networks

This paper presents a method to explain how input information is discard...

Generalisation and the Risk–Entropy Curve

In this paper we show that the expected generalisation performance of a ...

A short characterization of relative entropy

We prove characterization theorems for relative entropy (also known as K...

The edge of chaos: quantum field theory and deep neural networks

We explicitly construct the quantum field theory corresponding to a gene...

An Ode to an ODE

We present a new paradigm for Neural ODE algorithms, calledODEtoODE, whe...

Semiotic Aggregation in Deep Learning

Convolutional neural networks utilize a hierarchy of neural network laye...

Code Repositories


Code for project on relative entropy in deep neural networks

view repo