Gradient-based methods for spiking physical systems

08/29/2023

∙

Recent efforts have fostered significant progress towards deep learning in spiking networks, both theoretical and in silico. Here, we discuss several different approaches, including a tentative comparison of the results on BrainScaleS-2, and hint towards future such comparative studies.

READ FULL TEXT

Gradient-based methods for spiking physical systems

EventProp: Backpropagation for Exact Gradients in Spiking Neural Networks

Unifying Activation- and Timing-based Learning Rules for Spiking Neural Networks

Neurashed: A Phenomenological Model for Imitating Deep Learning Training

Gradient-based Neuromorphic Learning on Dynamical RRAM Arrays

Beyond backpropagation: implicit gradients for bilevel optimization

An Efficient Approach to Boosting Performance of Deep Spiking Network Training

Failures of Gradient-Based Deep Learning

Gradient-based methods for spiking physical systems

Related Research

EventProp: Backpropagation for Exact Gradients in Spiking Neural Networks

Unifying Activation- and Timing-based Learning Rules for Spiking Neural Networks

Neurashed: A Phenomenological Model for Imitating Deep Learning Training

Gradient-based Neuromorphic Learning on Dynamical RRAM Arrays

Beyond backpropagation: implicit gradients for bilevel optimization

An Efficient Approach to Boosting Performance of Deep Spiking Network Training

Failures of Gradient-Based Deep Learning