Neural ODE and Holographic QCD

Abstract

The neural ordinary differential equation (Neural ODE) is a novel machine learning architecture whose weights are smooth functions of the continuous depth. We apply the Neural ODE to holographic QCD by regarding the weight functions as a bulk metric, and train the machine with lattice QCD data of chiral condensate at finite temperature. The machine finds consistent bulk geometry at various values of temperature and discovers the emergent black hole horizon in the holographic bulk automatically. The holographic Wilson loops calculated with the emergent machine-learned bulk spacetime have consistent temperature dependence of confinement and Debye-screening behavior. In machine learning models with physically interpretable weights, the Neural ODE frees us from discretization artifact leading to difficult ingenuity of hyperparameters, and improves numerical accuracy to make the model more trustworthy.

10 pages, 6 figures

References in corpus (9)

• Machine learning and the physical sciences
• Exploring improved holographic theories for QCD: Part II
• Self-Learning Monte Carlo Method
• Entanglement Transitions from Holographic Random Tensor Networks
• Deep Learning and Holographic QCD
• AdS/CFT as a deep Boltzmann machine
• Machine Learning Holographic Mapping by Neural Network Renormalization Group
• Deep Quantum Geometry of Matrices
• QCD at non-zero temperature: bulk properties and heavy quarks

Cited by in corpus (15)

• Machine Learning in Nuclear Physics
• Exploring QCD matter in extreme conditions with Machine Learning
• Deriving dilaton potential in improved holographic QCD from meson spectrum
• Deep learning bulk spacetime from boundary optical conductivity
• Holographic reconstruction of black hole spacetime: machine learning and entanglement entropy
• Learning the black hole metric from holographic conductivity
• Disentangling the gravity dual of Yang-Mills theory
• Bulk reconstruction of metrics inside black holes by complexity
• Gravitational Duals from Equations of State
• Unification of Symmetries Inside Neural Networks: Transformer, Feedforward and Neural ODE
• Reconstructing black hole exteriors and interiors using entanglement and complexity
• Deep learning-based holography for T-linear resistivity
• The Algebraic Structure Underlying Pole-Skipping Points
• Capturing reduced-order quantum many-body dynamics out of equilibrium via neural ordinary differential equations
• Learning holographic QCD with unflavored meson spectra