Self-Assembly of a Biologically Plausible Learning Circuit

2024-12-28

AI Summary

The biological implausibility of backpropagation represents a fundamental tension between deep learning and neuroscience. Method: This paper proposes a biologically plausible learning framework grounded in self-organization, introducing heterosynaptic plasticity rules and random initial connectivity. Under an abstracted cortical ascending/descending pathway architecture, computational modeling spontaneously yields quadruplet synaptic loop structures. We further design, for the first time, a weight-update circuit exhibiting autonomous self-assembly—balancing neuroscientific plausibility with computational efficacy. Results: The framework achieves training performance comparable to backpropagation on standard benchmarks. Critically, it generates several experimentally testable predictions: anatomically specific synaptic localization, temporally structured plasticity dynamics, and precise circuit-level conditions governing self-organized loop formation. This work establishes a novel, empirically falsifiable paradigm for investigating how the brain learns.

Technology Category

• Machine Learning: Bio-inspired Learning
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Natural Language Processing: Learning & Optimization for NLP

Application Category

• Economics, Online Markets and Human Computation: Social networks and social learning
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

Over the last four decades, the amazing success of deep learning has been driven by the use of Stochastic Gradient Descent (SGD) as the main optimization technique. The default implementation for the computation of the gradient for SGD is backpropagation, which, with its variations, is used to this day in almost all computer implementations. From the perspective of neuroscientists, however, the consensus is that backpropagation is unlikely to be used by the brain. Though several alternatives have been discussed, none is so far supported by experimental evidence. Here we propose a circuit for updating the weights in a network that is biologically plausible, works as well as backpropagation, and leads to verifiable predictions about the anatomy and the physiology of a characteristic motif of four plastic synapses between ascending and descending cortical streams. A key prediction of our proposal is a surprising property of self-assembly of the basic circuit, emerging from initial random connectivity and heterosynaptic plasticity rules.

Problem

• Biologically-plausible learning
• Backpropagation alternative
• Brain-inspired computation

Innovation

• Biologically-inspired learning
• Connection strength update
• Automated circuit construction