Revealing systematics in phenomenologically viable flux vacua with reinforcement learning

Abstract

The organising principles underlying the structure of phenomenologically viable string vacua can be accessed by sampling such vacua. In many cases this is prohibited by the computational cost of standard sampling methods in the high dimensional model space. Here we show how this problem can be alleviated using reinforcement learning techniques to explore string flux vacua. We demonstrate in the case of the type IIB flux landscape that vacua with requirements on the expectation value of the superpotential and the string coupling can be sampled significantly faster by using reinforcement learning than by using metropolis or random sampling. Our analysis is on conifold and symmetric torus background geometries. We show that reinforcement learning is able to exploit successful strategies for identifying such phenomenologically interesting vacua. The strategies are interpretable and reveal previously unknown correlations in the flux landscape.

arXiv comment30 pages, 12 figures, 2 tables

References in corpus (27)

• PyTorch: An Imperative Style, High-Performance Deep Learning Library
• Dueling Network Architectures for Deep Reinforcement Learning
• Quantization of Four-form Fluxes and Dynamical Neutralization of the Cosmological Constant
• Flux Compactification
• The statistics of string/M theory vacua
• The Anthropic Landscape of String Theory
• Distributions of flux vacua
• Distributions of nonsupersymmetric flux vacua
• Flux Compactifications on Calabi-Yau Threefolds
• Computational complexity of the landscape I
• Enumerating Flux Vacua with Enhanced Symmetries
• Statistics on the Heterotic Landscape: Gauge Groups and Cosmological Constants of Four-Dimensional Heterotic Strings
• On the Explicit Construction and Statistics of Calabi-Yau Flux Vacua
• Critical points and supersymmetric vacua
• Low Energy Supersymmetry From the Landscape
• Supersymmetry Breaking in the Anthropic Landscape
• The landscape of intersecting brane models
• Statistics of M theory Vacua
• Branes with Brains: Exploring String Vacua with Deep Reinforcement Learning
• Branches of the Landscape
• On Algorithmic Universality in F-theory Compactifications
• Finding all flux vacua in an explicit example
• Scanning the skeleton of the 4D F-theory landscape
• Computational Complexity of Vacua and Near-Vacua in Field and String Theory
• Model Building and Phenomenology of Flux-Induced Supersymmetry Breaking on D3-branes
• GANs for generating EFT models
• Fast optimization algorithms and the cosmological constant

Cited by in corpus (9)

• Symmetries, Safety, and Self-Supervision
• Moduli Stabilization in Asymptotic Flux Compactifications
• Evolving Heterotic Gauge Backgrounds: Genetic Algorithms versus Reinforcement Learning
• Heterotic String Model Building with Monad Bundles and Reinforcement Learning
• Applying machine learning to the Calabi-Yau orientifolds with string vacua
• Breeding realistic D-brane models
• On the Search for Low $W_0$
• String Model Building, Reinforcement Learning and Genetic Algorithms
• Probing the Structure of String Theory Vacua with Genetic Algorithms and Reinforcement Learning