String Model Building, Reinforcement Learning and Genetic Algorithms

TL;DR

This paper compares reinforcement learning and genetic algorithms for efficiently discovering three-family models in heterotic Calabi-Yau compactifications, demonstrating their effectiveness in systematic searches where traditional methods are limited.

Contribution

It introduces the application of reinforcement learning and genetic algorithms to heterotic Calabi-Yau model building, showing their efficiency in finding phenomenologically relevant models.

Findings

Both methods efficiently identify three-family models.

They enable complete searches in complex model spaces.

Results reveal previously unknown models similar across methods.

Abstract

We investigate reinforcement learning and genetic algorithms in the context of heterotic Calabi-Yau models with monad bundles. Both methods are found to be highly efficient in identifying phenomenologically attractive three-family models, in cases where systematic scans are not feasible. For monads on the bi-cubic Calabi-Yau either method facilitates a complete search of the environment and leads to similar sets of previously unknown three-family models.