Evolving Heterotic Gauge Backgrounds: Genetic Algorithms versus Reinforcement Learning

TL;DR

This paper compares genetic algorithms and reinforcement learning as methods to navigate the string landscape, successfully identifying heterotic string compactifications that resemble the Standard Model of particle physics.

Contribution

It demonstrates that both genetic algorithms and reinforcement learning can effectively find anomaly-free heterotic string models with realistic features, offering new computational tools for string phenomenology.

Findings

Genetic algorithms successfully generated anomaly-free SO(10) GUTs with three families.

Reinforcement learning showed similar efficacy to genetic algorithms.

Both methods have complementary strengths in exploring the string landscape.

Abstract

The immensity of the string landscape and the difficulty of identifying solutions that match the observed features of particle physics have raised serious questions about the predictive power of string theory. Modern methods of optimisation and search can, however, significantly improve the prospects of constructing the standard model in string theory. In this paper we scrutinise a corner of the heterotic string landscape consisting of compactifications on Calabi-Yau three-folds with monad bundles and show that genetic algorithms can be successfully used to generate anomaly-free supersymmetric SO(10) GUTs with three families of fermions that have the right ingredients to accommodate the standard model. We compare this method with reinforcement learning and find that the two methods have similar efficacy but somewhat complementary characteristics.