Ultra-Peripheral Collisions as a Nuclear-Structure Interferometer with Interpretable Multitask Deep Learning
2026-06-22 • Machine Learning
Machine Learning
AI summaryⓘ
The authors explain a new way to study the tiny structure of atomic nuclei using special collisions called ultra-peripheral collisions (UPCs). These collisions produce patterns from light interacting with the nucleus that reveal its shape and density, but interpreting these patterns is complex. To solve this, the authors developed a deep-learning method that looks at these patterns and figures out multiple details about the nuclear structure all at once. They tested their method on a specific type of collision involving zirconium nuclei and showed it can separate different types of information for better understanding. This approach can help analyze future detailed data from such collisions.
Ultra-peripheral collisionsNuclear structureCoherent vector-meson photoproductionDiffraction patternsTwo-source interferenceDeep learningTransverse momentum distributionsNeutron skinNuclear deformationJ/ψ photoproduction
Authors
Jing-Zong Zhang, Wang-Mei Zha, Lingxiao Wang, Guo-Liang Ma
Abstract
Precise knowledge of nuclear structure is essential across fundamental physics, yet probing these structures is notoriously difficult. To address this challenge, ultra-peripheral collisions (UPCs) provide a femtoscopic tomography for imaging the atomic nucleus. UPCs offer a pristine electromagnetic pathway: coherent vector-meson photoproduction generates patterns of diffraction and two-source interference that directly encode the nuclear spatial density. Turning these patterns into quantitative constraints is, however, a challenging inverse problem, complicated by correlated sensitivities to deformation and neutron skin, phase smearing, and experimental backgrounds. Here we introduce an interpretable Multitask deep-learning framework that maps transverse momentum distributions to multiple nuclear-structure indicators simultaneously and identifies the kinematic regions driving each inference. We demonstrate the approach with coherent $J/ψ$ photoproduction in $^{96}_{40}\text{Zr} + ^{96}_{40}\text{Zr}$ collisions, showing that the learned features separate diffraction-dominated and interference-dominated information and provide analysis-ready observables for future high-luminosity data.