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

TL;DR

This paper develops a quantum field theory framework for deep neural networks, analyzing criticality and fluctuations to understand information propagation and trainability.

Contribution

It constructs a QFT model for neural networks, deriving criticality conditions and loop corrections, linking neural dynamics to well-known physical models.

Findings

Derived the mean-field theory and criticality conditions for neural networks.

Computed loop corrections analogous to the $O(N)$ vector model.

Identified how finite-width effects influence information propagation.

Abstract

We explicitly construct the quantum field theory corresponding to a general class of deep neural networks encompassing both recurrent and feedforward architectures. We first consider the mean-field theory (MFT) obtained as the leading saddlepoint in the action, and derive the condition for criticality via the largest Lyapunov exponent. We then compute the loop corrections to the correlation function in a perturbative expansion in the ratio of depth $T$ to width $N$, and find a precise analogy with the well-studied $O(N)$ vector model, in which the variance of the weight initializations plays the role of the 't Hooft coupling. In particular, we compute both the $\mathcal{O}(1)$ corrections quantifying fluctuations from typicality in the ensemble of networks, and the subleading $\mathcal{O}(T/N)$ corrections due to finite-width effects. These provide corrections to the correlation length that controls the depth to which information can propagate through the network, and thereby sets the scale at which such networks are trainable by gradient descent. Our analysis provides a first-principles approach to the rapidly emerging NN-QFT correspondence, and opens several interesting avenues to the study of criticality in deep neural networks.