Orateur
Description
Understanding the 3D structure of visible matter in the universe while advancing the tomography of nucleons is one of the central goals of contemporary nuclear physics, pursued at the upgraded 12 GeV Jefferson Lab accelerator and at the planned Electron-Ion Collider (EIC). To fully capitalize on the data emerging from these experiments and to guide their extraction, the DOE-funded EXCLusives via Artificial Intelligence and Machine learning (EXCLAIM) collaboration is developing a framework for the next-generation precision characterization of the quark-gluon structure of matter. Central to our initiative is the design of physics informed deep learning architectures that, unlike standard “black box” methods, are crafted for specific physics purposes using more flexible and physically explainable ML building blocks. This program requires a joint effort of ML experts and nuclear theorists/experimentalists to work together in interdisciplinary collaborative work. Our approach to 3D hadron structure aims at providing the technology to continually incorporate both the latest experimental data and precision first-principle Lattice QCD (LQCD) calculations. We will present a benchmark system to track the behavior of various asymmetries for Deeply Virtual Exclusive Scattering (DVES) processes, enlarging the scope of the tomographic explorations to include processes characterized by multi-particle final states many-body correlations.
