Modern Machine Learning and Particle Physics Phenomenology at the LHC

TL;DR

The paper addresses the challenge of delivering accurate, efficient, and probabilistic predictions for LHC phenomenology across the entire prediction pipeline, from parton-level calculations to end-to-end simulations. It surveys ML strategies that surrogate key ingredients such as amplitudes, phase-space integration, PDFs, and SMEFT parameter inference, and extends to end-to-end event simulation with diffusion models and normalising flows, all while emphasising uncertainty quantification and symmetry-aware design. Notable contributions include ML surrogates for $|M|^2$ and $d\sigma/d\Phi$, the MadNIS framework, NNPDF4.0 with bootstrap replicas, SMEFT likelihood learning, OTUS, and MEM-inspired ML integrations, which together accelerate and improve precision of collider predictions. The work highlights the potential for probabilistic, interpretable, and symmetry-respecting ML methodologies to enhance both SM and beyond-SM physics analyses at the LHC.

Abstract

Modern machine learning is driving a paradigm shift in particle physics phenomenology at the Large Hadron Collider. This short review examines the transformative role of machine learning across the entire theoretical prediction pipeline, from parton-level calculations to full simulations. We discuss applications to scattering amplitude computations, phase space integration, Parton Distribution Function determination, and parameter extraction. Some critical frontiers for the field including uncertainty quantification, the role of symmetries, and interpretability are highlighted.