Neural network extraction of chromo-electric and chromo-magnetic gluon masses

TL;DR

This paper introduces a neural network model that accurately extracts temperature-dependent chromo-electric and chromo-magnetic gluon masses from lattice QCD data, revealing their distinct behaviors across the deconfinement transition.

Contribution

It develops a novel dual residual neural network approach with physics regularizations to separate and analyze gluon mass contributions from lattice thermodynamics data.

Findings

High-accuracy reproduction of lattice results across temperature range

Identification of sharp mass decrease near critical temperature

Observation of convergence of masses at high temperatures

Abstract

We present a neural network-based quasi-particle model to separate the contributions of chromo-electric and chromo-magnetic gluons. Using dual residual networks, we extract temperature-dependent masses from SU(3) lattice thermodynamic data of pressure and trace anomaly. After incorporating physics regularizations, the trained models reproduce lattice results with high accuracy over $T/T_c \in [1,10]$, capturing both the crossover behavior near $T_c$ and linear scaling at high temperatures. The extracted masses exhibit a physically reasonable behavior: they decrease sharply around $T_c$ and increase linearly thereafter. We find significant differences between thermal and screening masses near $T_c$, reflecting non-perturbative dynamics, while they converge at $T \gtrsim 2T_c$.