3:30 pm
MCP 201 933 East 56th Street
Chicago IL 60637
Modern machine learning is revolutionizing the way we explore the universe — from the smallest particles to the largest structures in the cosmos. At the Large Hadron Collider, a major effort is underway to reimagine how searches for new physics can be performed in a more model-agnostic manner, using powerful new data-driven anomaly detection methods. In this talk, I will give an overview of this effort, showing how the same building blocks that give rise to today’s generative AI models can be repurposed to uncover hidden patterns in particle collisions. Remarkably, the methods we are developing also cross disciplinary boundaries -- I will show how they can also be applied to astronomical data to discover stellar streams, reveal ultra-faint dwarf galaxies, and even map the local gravitational potential and dark matter density of our Galaxy. Together, these advances suggest a new era where machine learning becomes not just a tool, but a discovery engine for fundamental physics.