Desynchronous Learning in a Physics-Driven Learning Network

TL;DR

This paper explores desynchronous learning in physics-driven neural networks, demonstrating that it maintains or improves performance and enhances scalability compared to traditional synchronized updates.

Contribution

It introduces and validates the concept of desynchronous learning in decentralized physics-driven networks, showing its benefits over synchronized methods.

Findings

Desynchronization does not degrade performance in simulations.

Desynchronization improves exploration of solution space.

Analogous effects of desynchronization and mini-batching are observed.

Abstract

In a neuron network, synapses update individually using local information, allowing for entirely decentralized learning. In contrast, elements in an artificial neural network (ANN) are typically updated simultaneously using a central processor. Here we investigate the feasibility and effect of desynchronous learning in a recently introduced decentralized, physics-driven learning network. We show that desynchronizing the learning process does not degrade performance for a variety of tasks in an idealized simulation. In experiment, desynchronization actually improves performance by allowing the system to better explore the discretized state space of solutions. We draw an analogy between desynchronization and mini-batching in stochastic gradient descent, and show that they have similar effects on the learning process. Desynchronizing the learning process establishes physics-driven learning networks as truly fully distributed learning machines, promoting better performance and scalability in deployment.