Decoding high energy physics with AI and machine learning
Adobe Stock
Artificial intelligence is transforming the way scientists explore the fundamental forces of the universe—and a University of Arizona researcher is playing a key role. In a recent collaboration with the U.S. Department of Energy’s Argonne National Laboratory, physicists unveiled new AI-based tools that are enhancing our understanding of parton distribution functions (PDFs), which describe the behavior of quarks and gluons inside protons.
These tools, called PDFdecoder and XAI4PDF, use machine learning models to improve both the accuracy and interpretability of particle physics data. PDFdecoder employs generative AI to reconstruct missing details about particle behavior, while XAI4PDF uses explainable AI techniques to make the decision-making processes of AI models more transparent. Together, these frameworks help bridge theoretical physics with experimental results from high-energy particle colliders.
Jonathon Gomprecht, a researcher from the University of Arizona, contributed to this groundbreaking work. By integrating expertise in AI and theoretical physics, this research strengthens the connection between scientific modeling and real-world data, potentially paving the way for new discoveries beyond the Standard Model of particle physics.