Stochastic dendrites enable online learning in mixed-signal neuromorphic processing systems

TL;DR

This paper introduces a spike-based online learning method inspired by dendritic compartments in neurons, enabling efficient, robust, and low-precision learning in neuromorphic systems for edge computing.

Contribution

It presents a novel stochastic dendritic learning circuit design implemented in a 180nm neuromorphic chip, demonstrating effective online learning with 4-bit weights on MNIST.

Findings

Achieved over 85% accuracy on MNIST with 4-bit weights.

Demonstrated robust online learning in a prototype neuromorphic system.

Validated the circuit and algorithm co-design approach through behavioral simulations.

Abstract

The stringent memory and power constraints required in edge-computing sensory-processing applications have made event-driven neuromorphic systems a promising technology. On-chip online learning provides such systems the ability to learn the statistics of the incoming data and to adapt to their changes. Implementing online learning on event driven-neuromorphic systems requires (i) a spike-based learning algorithm that calculates the weight updates using only local information from streaming data, (ii) mapping these weight updates onto limited bit precision memory and (iii) doing so in a robust manner that does not lead to unnecessary updates as the system is reaching its optimal output. Recent neuroscience studies have shown how dendritic compartments of cortical neurons can solve these problems in biological neural networks. Inspired by these studies we propose spike-based learning circuits to implement stochastic dendritic online learning. The circuits are embedded in a prototype spiking neural network fabricated using a 180nm process. Following an algorithm-circuits co-design approach we present circuits and behavioral simulation results that demonstrate the learning rule features. We validate the proposed method using behavioral simulations of a single-layer network with 4-bit precision weights applied to the MNIST benchmark and demonstrating results that reach accuracy levels above 85%.