Distributed Representations Enable Robust Multi-Timescale Symbolic Computation in Neuromorphic Hardware

Abstract

Programming recurrent spiking neural networks (RSNNs) to robustly perform multi-timescale computation remains a difficult challenge. To address this, we describe a single-shot weight learning scheme to embed robust multi-timescale dynamics into attractor-based RSNNs, by exploiting the properties of high-dimensional distributed representations. We embed finite state machines into the RSNN dynamics by superimposing a symmetric autoassociative weight matrix and asymmetric transition terms, which are each formed by the vector binding of an input and heteroassociative outer-products between states. Our approach is validated through simulations with highly nonideal weights; an experimental closed-loop memristive hardware setup; and on Loihi 2, where it scales seamlessly to large state machines. This work introduces a scalable approach to embed robust symbolic computation through recurrent dynamics into neuromorphic hardware, without requiring parameter fine-tuning or significant platform-specific optimisation. Moreover, it demonstrates that distributed symbolic representations serve as a highly capable representation-invariant language for cognitive algorithms in neuromorphic hardware.

19 pages, 7 figures. Supplementary material: 13 pages, 8 figures. Accepted for publication in Neuromorphic Computing and Engineering

References in corpus (14)

• Weak pairwise correlations imply strongly correlated network states in a neural population
• Neuromorphic electronic circuits for building autonomous cognitive systems
• A Survey on Hyperdimensional Computing aka Vector Symbolic Architectures, Part I: Models and Data Transformations
• Neuromorphic Hardware In The Loop: Training a Deep Spiking Network on the BrainScaleS Wafer-Scale System
• A Theoretical Perspective on Hyperdimensional Computing
• State dependent computation using coupled recurrent networks
• Vector Symbolic Architectures as a Computing Framework for Emerging Hardware
• Neuromorphic Intermediate Representation: A Unified Instruction Set for Interoperable Brain-Inspired Computing
• Neuromorphic Visual Scene Understanding with Resonator Networks
• Rhythmic inhibition allows neural networks to search for maximally consistent states
• Scaling Limits of Memristor-Based Routers for Asynchronous Neuromorphic Systems
• Vector Symbolic Finite State Machines in Attractor Neural Networks
• Visual Pattern Recognition with on On-chip Learning: towards a Fully Neuromorphic Approach
• Gradient-descent hardware-aware training and deployment for mixed-signal Neuromorphic processors