Abide by the Law and Follow the Flow: Conservation Laws for Gradient Flows

Abstract

Understanding the geometric properties of gradient descent dynamics is a key ingredient in deciphering the recent success of very large machine learning models. A striking observation is that trained over-parameterized models retain some properties of the optimization initialization. This "implicit bias" is believed to be responsible for some favorable properties of the trained models and could explain their good generalization properties. The purpose of this article is threefold. First, we rigorously expose the definition and basic properties of "conservation laws", that define quantities conserved during gradient flows of a given model (e.g. of a ReLU network with a given architecture) with any training data and any loss. Then we explain how to find the maximal number of independent conservation laws by performing finite-dimensional algebraic manipulations on the Lie algebra generated by the Jacobian of the model. Finally, we provide algorithms to: a) compute a family of polynomial laws; b) compute the maximal number of (not necessarily polynomial) independent conservation laws. We provide showcase examples that we fully work out theoretically. Besides, applying the two algorithms confirms for a number of ReLU network architectures that all known laws are recovered by the algorithm, and that there are no other independent laws. Such computational tools pave the way to understanding desirable properties of optimization initialization in large machine learning models.

References in corpus (11)

• Reconciling modern machine learning practice and the bias-variance trade-off
• Exact solutions to the nonlinear dynamics of learning in deep linear neural networks
• In Search of the Real Inductive Bias: On the Role of Implicit Regularization in Deep Learning
• A Convergence Analysis of Gradient Descent for Deep Linear Neural Networks
• Implicit Regularization in Deep Learning
• Implicit Bias of Gradient Descent for Wide Two-layer Neural Networks Trained with the Logistic Loss
• On the Implicit Bias of Initialization Shape: Beyond Infinitesimal Mirror Descent
• Noether: The More Things Change, the More Stay the Same
• Gradient descent follows the regularization path for general losses
• Symmetries, flat minima, and the conserved quantities of gradient flow
• Implicit Bias of Gradient Descent on Reparametrized Models: On Equivalence to Mirror Descent