Speaker
Description
Extracting the quark and gluon structure of hadrons from high-energy scattering data is a challenging inverse problem that requires flexible representations, efficient QCD calculations, and reliable uncertainty quantification. I will discuss recent progress toward a computational framework that connects measured events directly to the underlying partonic correlation functions. A central component is the event-level analysis being developed within the SciDAC QuantOm project, which combines theoretical calculations with realistic experimental conditions in a scalable inference workflow. I will then describe complementary advances in reconstructing hadron structure, including finite-element-based methods for generalized parton distributions and pixel-based Bayesian imaging of transverse-momentum-dependent distributions using generative AI. Finally, I will discuss ongoing extensions of these ideas to collinear parton distribution functions and fragmentation functions within the JAM global-analysis framework. Together, these developments point toward a unified approach that uses modern computational and machine-learning tools to extract maximal information from current experiments and future measurements at the Electron–Ion Collider.