Fault-Tolerant Neural Networks from Biological Error Correction Codes

TL;DR

This paper demonstrates that biological error correction codes enable neural networks to perform reliable computation despite neuron unreliability, revealing a phase transition threshold and offering insights into cortical and artificial neural systems.

Contribution

It introduces a universal fault-tolerant neural network model based on biological error correction codes, establishing a threshold for neuron noise tolerance and demonstrating fault-tolerant computation in neural systems.

Findings

Neural networks can achieve reliable computation below a specific noise threshold.

A phase transition from faulty to fault-tolerant computation was identified.

Biological neurons can operate reliably despite inherent noise levels.

Abstract

It has been an open question in deep learning if fault-tolerant computation is possible: can arbitrarily reliable computation be achieved using only unreliable neurons? In the grid cells of the mammalian cortex, analog error correction codes have been observed to protect states against neural spiking noise, but their role in information processing is unclear. Here, we use these biological error correction codes to develop a universal fault-tolerant neural network that achieves reliable computation if the faultiness of each neuron lies below a sharp threshold; remarkably, we find that noisy biological neurons fall below this threshold. The discovery of a phase transition from faulty to fault-tolerant neural computation suggests a mechanism for reliable computation in the cortex and opens a path towards understanding noisy analog systems relevant to artificial intelligence and neuromorphic computing.