Interpretable deep learning for nuclear deformation in heavy ion collisions

Abstract

The structure of heavy nuclei is difficult to disentangle in high-energy heavy-ion collisions. The deep convolution neural network (DCNN) might be helpful in mapping the complex final states of heavy-ion collisions to the nuclear structure in the initial state. Using DCNN for supervised regression, we successfully extracted the magnitude of the nuclear deformation from event-by-event correlation between the momentum anisotropy or elliptic flow ($v_2$) and total number of charged hadrons ($d{N}_{\mathrm{ch}}/d\eta$) within a Monte Carlo model. Furthermore, a degeneracy is found in the correlation between collisions of prolate-prolate and oblate-oblate nuclei. Using the Regression Attention Mask algorithm which is designed to interpret what has been learned by DCNN, we discovered that the correlation in total-overlapped collisions is sensitive to only large nuclear deformation, while the correlation in semi-overlapped collisions is discriminative for all magnitudes of nuclear deformation. The method developed in this study can pave a way for exploration of other aspects of nuclear structure in heavy-ion collisions.

8 pages, 2 figures, AI + X research

References in corpus (25)

• Towards A Rigorous Science of Interpretable Machine Learning
• Axiomatic Attribution for Deep Networks
• Striving for Simplicity: The All Convolutional Net
• Squeeze-and-Excitation Networks
• Computing Quark and Gluon Distribution Functions for Very Large Nuclei
• Predicting Cardiovascular Risk Factors from Retinal Fundus Photographs using Deep Learning
• Glauber Modeling in High Energy Nuclear Collisions
• Gluon Distribution Functions for Very Large Nuclei at Small Transverse Momentum
• Explaining NonLinear Classification Decisions with Deep Taylor Decomposition
• Alternative ansatz to wounded nucleon and binary collision scaling in high-energy nuclear collisions
• Visualizing Deep Neural Network Decisions: Prediction Difference Analysis
• Event-by-event gluon multiplicity, energy density and eccentricities at RHIC and LHC
• Mapping the hydrodynamic response to the initial geometry in heavy-ion collisions
• Learning how to explain neural networks: PatternNet and PatternAttribution
• Pseudorapidity distribution and decorrelation of anisotropic flow within CLVisc hydrodynamics
• The (Un)reliability of saliency methods
• Testing the Chiral Magnetic Effect with Central U+U collisions
• Azimuthal anisotropy in U+U and Au+Au collisions at RHIC
• Linear and cubic response to the initial eccentricity in heavy-ion collisions
• Collision Geometry and Flow in Uranium+Uranium Collisions
• Multiplicity distribution and source deformation in full-overlap U+U collisions
• Signatures of $\alpha$ clustering in ultra-relativistic collisions with light nuclei
• Massively parallel simulations of relativistic fluid dynamics on graphics processing units with CUDA
• Initial state anisotropies and their uncertainties in ultrarelativistic heavy-ion collisions from the Monte Carlo Glauber model
• Effect of initial-state nucleon-nucleon correlations on collective flow in ultra-central heavy-ion collisions

Cited by in corpus (4)

• Deep learning stochastic processes with QCD phase transition
• Explainable machine learning of the underlying physics of high-energy particle collisions
• Applications of deep learning to relativistic hydrodynamics
• Report from the A.I. For Nuclear Physics Workshop