Beyond Images: Leveraging Stable Diffusion Techniques for Particle Physics Simulations

by Prof. Eilam Gross

WIS

at Physics Colloquium

Tue, 28 Oct 2025, 16:00
Ilse Katz Institute for Nanoscale Science & Technology (51), room 015

Abstract

Generative AI models, including those behind image creation tools, have shown remarkable capabilities in transforming random inputs into coherent outputs. Inspired by these advancements, we've developed Parnassus, a deep-learning model designed for particle physics. Parnassus processes point clouds representing particles interacting with a detector and outputs reconstructed particle data. Parnassus accurately replicates the particle flow algorithm and generalizes well beyond its training set. This approach exemplifies how techniques from image generation can be adapted to accelerate simulations in high-energy physics.

Created on 21-10-2025 by Bar Lev, Yevgeny (ybarlev)
Updaded on 21-10-2025 by Bar Lev, Yevgeny (ybarlev)