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.

Deep neural networks: a new framework for modelling biological vision and brain information processing

Neural networks and neuroscience-inspired computer vision

Convolutional neural networks for vision neuroscience: significance, developments, and outstanding issues

A deep learning framework for neuroscience

Deep Learning for Cognitive Neuroscience

Neuroscientific insights about computer vision models: a concise review

Exploring the Brain-like Properties of Deep Neural Networks: A Neural Encoding Perspective

Towards Human-Leveled Vision Systems

Neural network models and deep learning - a primer for biologists

Seeing eye-to-eye? A comparison of object recognition performance in humans and deep convolutional neural networks under image manipulation

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

Deep Neural Networks Rival the Representation of Primate IT Cortex for Core Visual Object Recognition

From deep learning to mechanistic understanding in neuroscience: the structure of retinal prediction

Unidirectional brain-computer interface: Artificial neural network encoding natural images to fMRI response in the visual cortex

Neural Networks, Artificial Intelligence and the Computational Brain

Brain-inspired Models for Visual Object Recognition: an Overview

Deep Learning for Visual Neuroprosthesis

A Biologically Inspired Neurocomputing Circuit for Image Representation

A Bio-inspired Model for Image Representation and Image Analysis

Neural Dynamics Model of Visual Decision-Making: Learning from Human Experts

Canonical circuit computations for computer vision