Generative inverse mappers for high energy physics -- Brandon Kriesten (Argonne)

​We’re excited to announce our next online event featuring a talk by Brandon Kriesten from Argonne National Laboratory! Join us for an engaging session on how generative AI and evidential learning are reshaping our understanding of hadron structure and uncertainty quantification in high-energy physics.

​❗ Hosted by AI for HEP (join the community) on alphaXiv , the recording can be found on alphaXiv.

​📅 Wednesday October 15th, 2025 · 10AM PT

​🎙️ Featuring Brandon Kriesten (Argonne National Laboratory, HEP Theory Group)

​💬 30-min Talk + Interactive Discussion

​Brandon will discuss how generative AI approaches can be used to tackle inverse problems in hadron structure and tomography—where precisely capturing uncertainty is key to uncovering new physics.

​By bridging physics and next-generation AI/ML frameworks, this work leverages evidential deep learning and information-theoretic metrics to disentangle aleatoric, epistemic, and distributional uncertainties.

​Through integrating lattice QCD constraints and experimental data, these models enable refined extractions of parton distribution functions (PDFs) and open new avenues for probing nonperturbative QCD and beyond-the-Standard-Model phenomena.

​This talk will highlight how physics-informed machine learning not only advances our understanding of hadron structure but also pushes the frontier of precision QCD in the AI era.