Gradient Flow Equations for Deep Linear Neural Networks: A Survey from a Network Perspective

TL;DR

This survey reviews recent advances in understanding the dynamics and loss landscape of gradient flow equations in deep linear neural networks, highlighting their mathematical properties and critical point structure.

Contribution

It provides a comprehensive analysis of the gradient flow equations formulated as matrix ODEs, revealing their nilpotent, polynomial, and isospectral nature, and describes the loss landscape's critical points and invariants.

Findings

Gradient flow equations form nilpotent, polynomial, isospectral matrix ODEs.

Loss landscape has infinitely many global minima and saddle points, no local minima.

Critical values correspond to singular values of data learned by the network.

Abstract

The paper surveys recent progresses in understanding the dynamics and loss landscape of the gradient flow equations associated to deep linear neural networks, i.e., the gradient descent training dynamics (in the limit when the step size goes to 0) of deep neural networks missing the activation functions and subject to quadratic loss functions. When formulated in terms of the adjacency matrix of the neural network, as we do in the paper, these gradient flow equations form a class of converging matrix ODEs which is nilpotent, polynomial, isospectral, and with conservation laws. The loss landscape is described in detail. It is characterized by infinitely many global minima and saddle points, both strict and nonstrict, but lacks local minima and maxima. The loss function itself is a positive semidefinite Lyapunov function for the gradient flow, and its level sets are unbounded invariant sets of critical points, with critical values that correspond to the amount of singular values of the input-output data learnt by the gradient along a certain trajectory. The adjacency matrix representation we use in the paper allows to highlight the existence of a quotient space structure in which each critical value of the loss function is represented only once, while all other critical points with the same critical value belong to the fiber associated to the quotient space. It also allows to easily determine stable and unstable submanifolds at the saddle points, even when the Hessian fails to obtain them.