Abstract:Deep neural feedforward networks are effective models for a wide array of problems, but training and deploying such networks presents a significant energy cost. Spiking neural networks (SNNs), which are modeled after biologically realistic neurons, offer a potential solution when deployed correctly on neuromorphic computing hardware. Still, many applications train SNNs offline, and running network training directly on neuromorphic hardware is an ongoing research problem. The primary hurdle is that back-propagation, which makes training such artificial deep networks possible, is biologically implausible. Neuroscientists are uncertain about how the brain would propagate a precise error signal backward through a network of neurons. Recent progress addresses part of this question, e.g., the weight transport problem, but a complete solution remains intangible. In contrast, novel learning rules based on the information bottleneck (IB) train each layer of a network independently, circumventing the need to propagate errors across layers. Instead, propagation is implicit due the layers' feedforward connectivity. These rules take the form of a three-factor Hebbian update a global error signal modulates local synaptic updates within each layer. Unfortunately, the global signal for a given layer requires processing multiple samples concurrently, and the brain only sees a single sample at a time. We propose a new three-factor update rule where the global signal correctly captures information across samples via an auxiliary memory network. The auxiliary network can be trained a priori independently of the dataset being used with the primary network. We demonstrate comparable performance to baselines on image classification tasks. Interestingly, unlike back-propagation-like schemes where there is no link between learning and memory, our rule presents a direct connection between working memory and synaptic updates. To the best of our knowledge, this is the first rule to make this link explicit. We explore these implications in initial experiments examining the effect of memory capacity on learning performance. Moving forward, this work suggests an alternate view of learning where each layer balances memory-informed compression against task performance. This view naturally encompasses several key aspects of neural computation, including memory, efficiency, and locality.

Memory-enriched computation and learning in spiking neural networks through Hebbian plasticity

Memory-Dependent Computation and Learning in Spiking Neural Networks Through Hebbian Plasticity

A Brain-Inspired Spiking Neural Network Model with Temporal Encoding and Learning

Toward Efficient Processing and Learning with Spikes: New Approaches for Multispike Learning

A Hierarchical Organized Memory Model Using Spiking Neurons

A history of spike-timing-dependent plasticity

Spiking representation learning for associative memories

Optimality of short-term synaptic plasticity in modelling certain dynamic environments

Information bottleneck-based Hebbian learning rule naturally ties working memory and synaptic updates

Stable Hebbian Learning from Spike Timing-Dependent Plasticity

Emergence and maintenance of modularity in neural networks with Hebbian and anti-Hebbian inhibitory STDP

Beyond spiking networks: The computational advantages of dendritic amplification and input segregation

A memory neural system built based on spiking neural network

Computing of temporal information in spiking neural networks with ReRAM synapses

Experimental Demonstration of Supervised Learning in Spiking Neural Networks with Phase-Change Memory Synapses

Learning to learn with backpropagation of Hebbian plasticity

Learning the Plasticity: Plasticity-Driven Learning Framework in Spiking Neural Networks

Implementation of Memristive Neural Networks with Spike-Rate-dependent Plasticity Synapses

Building a realistic, scalable memory model with independent engrams using a homeostatic mechanism

Towards deep learning with spiking neurons in energy based models with contrastive Hebbian plasticity