Abstract:Understanding the computational principles of the brain and replicating them on neuromorphic hardware and modern deep learning architectures is crucial for advancing neuro-inspired AI (NeuroAI). Here, we develop an experimentally-constrained biophysical network model of neocortical circuit motifs, focusing on layers 2-3 of the primary visual cortex (V1). We investigate the role of four major cortical interneuron classes in a competitive-cooperative computational primitive and validate these circuit motifs implemented soft winner-take-all (sWTA) computation for gain modulation, signal restoration, and context-dependent multistability. Using a novel parameter mapping technique, we configured IBM's TrueNorth (TN) chip to implement sWTA computations, mirroring biological neural dynamics. Retrospectively, we observed a strong correspondence between the biophysical model and the TN hardware parameters, particularly in the roles of four key inhibitory neuron classes: Parvalbumin (feedforward inhibition), Somatostatin (feedback inhibition), VIP (disinhibition), and LAMP5 (gain normalization). Moreover, the sparse coupling of this sWTA motif was also able to simulate a two-state neural state machine on the TN chip, replicating working memory dynamics essential for cognitive tasks. Additionally, integrating the sWTA computation as a pre-processing layer in the Vision Transformer (ViT) enhanced its performance on the MNIST digit classification task, demonstrating improved generalization to previously unseen data and suggesting a mechanism akin to zero-shot learning. Our approach provides a framework for translating brain-inspired computations to neuromorphic hardware, with potential applications on platforms like Intel's Loihi2 and IBM's Northpole. By integrating biophysically accurate models with neuromorphic hardware and advanced machine learning techniques, we offer a comprehensive roadmap for embedding neural computation into NeuroAI systems.

Recognition of environment for navigation of gas pipeline inspection robot based on monocular vision system

Mapping Very Large Scale Spiking Neuron Network to Neuromorphic Hardware.

Efficient Point-Process Modeling of Spiking Neurons for Neuroprosthesis

Emerging Devices for Biologically Accurate Neuron

A soma‐synapses neuron model and FPGA implementation

A Spiking Visual Neuron for Depth Perceptual Systems

An All-Neuron Spiking 2-D Path Integration and Map Representation Model Implemented on Neuromorphic Chips

Real-time multicompartment Hodgkin-Huxley neuron emulation on SoC FPGA

A Neuromorphic-Hardware Oriented Bio-Plausible Online-Learning Spiking Neural Network Model.

An Artificial Visual Neuron with Multiplexed Rate and Time-to-first-spike Coding

Homogeneous Spiking Neuromorphic System for Real-World Pattern Recognition

A Novel Nonlinear Function Evaluation Approach for Efficient FPGA Mapping of Neuron and Synaptic Plasticity Models

Design and Implementation of a Primary Visual Cortex Pathway Model Based on Opponent-process Theory

Biologically Realistic Computational Primitives of Neocortex Implemented on Neuromorphic Hardware Improve Vision Transformer Performance

Modular Building Blocks for Mapping Spiking Neural Networks Onto a Programmable Neuromorphic Processor

A calibratable sensory neuron based on epitaxial VO2 for spike-based neuromorphic multisensory system

A FPGA-Based, Granularity-Variable Neuromorphic Processor and Its Application in a MIMO Real-Time Control System.

A Versatile Neuromorphic System Based on Simple Neuron Model

A 5.37-TSOPS/W Reconfigurable Neuron Array with Dual-mode Neurons and Asynchronous Synapses for Energy-Efficient Inference and Biological Neural Network Simulation

A Multichip Neuromorphic System for Spike-Based Visual Information Processing

Optoelectronic neuron based on transistor combined with volatile threshold switching memristors for neuromorphic computing