Erlangen 2026 – scientific programme

Parts | Days | Selection | Search | Updates | Downloads | Help

T: Fachverband Teilchenphysik

T 64: Invited Topical Talks III

T 64.4: Invited Topical Talk

Thursday, March 19, 2026, 15:15–15:45, AudiMax

Modern Machine Learning for LHC Event Generation — •Ramon Winterhalder — TIFLab, Università degli Studi di Milano & INFN Sezione di Milano, Italy

High-precision simulations from first principles are a cornerstone of LHC physics. With the upcoming high-luminosity phase of the LHC and the significant increase in experimental data, traditional simulation pipelines face growing challenges in terms of computational cost and efficiency. In this talk, I will discuss how modern machine-learning methods and computing paradigms can accelerate event generation while preserving theoretical accuracy. I will focus on three distinct and complementary approaches: neural importance sampling, as implemented in the MadNIS framework; machine-learned surrogate models for expensive matrix-element calculations; and hardware-aware implementations that exploit GPU acceleration and parallelization for efficient large-scale deployment. Together, these developments significantly reduce computational cost and pave the way towards the first ML-based event generator.

Keywords: Event Generation; MadGraph; Machine Learning; Monte Carlo Simulation