Probing criticality with deep learning in relativistic heavy-ion collisions

TL;DR

This paper demonstrates that deep learning can identify critical fluctuations in heavy-ion collision data by embedding and detecting signals modeled on the 3D Ising universality class, revealing subtle correlations.

Contribution

It introduces a point cloud neural network with dynamical edge convolution to detect critical phenomena in heavy-ion collision simulations, leveraging universality principles.

Findings

Successfully identifies critical fluctuations in simulated data.

Detects a significant fraction of signal particles in each event.

Shows deep learning can uncover subtle critical signals within complex particle clouds.

Abstract

Systems with different interactions could develop the same critical behaviour due to the underlying symmetry and universality. Using this principle of universality, we can embed critical correlations modeled on the 3D Ising model into the simulated data of heavy-ion collisions, hiding weak signals of a few inter-particle correlations within a large particle cloud. Employing a point cloud network with dynamical edge convolution, we are able to identify events with critical fluctuations through supervised learning, and pick out a large fraction of signal particles used for decision-making in each single event.