Desynchronous Learning in a Physics-Driven Learning Network
This addresses the problem of decentralized learning for scalable AI systems, though it appears incremental as it builds on existing physics-driven networks.
The paper investigated desynchronous learning in a physics-driven network, showing that it does not degrade performance in simulations and actually improves it in experiments by enhancing exploration of solution spaces.
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.