String Theory from Infinite Width Neural Networks

TL;DR

This paper demonstrates how infinite width neural networks can model bosonic string theory, enabling new computations of fundamental amplitudes through neural network correlators.

Contribution

It introduces a novel connection between neural networks and string theory, providing a new framework for computing string amplitudes using neural network ensembles.

Findings

Neural networks can realize bosonic string theory.

String tension is controlled by output weight variance.

Virasoro-Shapiro and Veneziano amplitudes are computed as neural correlators.

Abstract

We realize bosonic string theory with ensembles of infinite width neural networks. The string tension is tuned by the variance of the output weights. The construction provides a new computation of the foundational Virasoro-Shapiro and Veneziano amplitudes as neural network correlators.