Colloquium – Taoli Cheng – Towards AI-assisted High Energy Physics: A New Paradigm of Understanding Elementary Particles

Speaker: Taoli Cheng (University of Montreal)

Title: Towards AI-assisted High Energy Physics: A New Paradigm of Understanding Elementary Particles

Abstract:

The rapid evolution of artificial intelligence (AI) and deep learning methodologies is driving a profound paradigm shift in scientific discovery. Armed with large-scale datasets and the ability to handle high dimensionality, deep neural networks offer innovative approaches to tackle conventional physics challenges, including event simulation, particle identification, and signal detection. In this talk, I will first discuss my research efforts in developing deep learning models customized for analyzing physics events and automating scientific discovery in high energy physics. Taking a leap forward in building a unified and robust AI system, I will introduce my recent work on a novel energy-based probabilistic modeling framework. This framework aims to precisely describe the radiation patterns of elementary particles at the Large Hadron Collider. The learned energy function exhibits remarkable efficacy in addressing multiple pivotal physics tasks, especially in event simulation and model-independent searches for beyond-standard-model physics. I conclude by outlining my research vision for building advanced, controllable, and robust large-scale AI systems for driving scientific discovery.

Host: Sergei Gleyzer