Learning Standard Model structure from LHC data with Riemannian flow matching
2026-07-17 • Machine Learning
Machine Learning
AI summaryⓘ
The authors show that a single AI model, based on transformers, can learn details of particle physics from real collision data spanning a huge range of energies. Their model, called ShellFlow, uses just basic physics rules about particles and masses to generate realistic particles seen in experiments. Training on a billion events from ATLAS data, the model accurately recreates known particle properties like particle masses and angles without being explicitly told about them. This suggests important parts of the Standard Model can be learned directly from data alone.
Transformer modelGenerative modelStandard ModelInvariant massMonte Carlo simulationRiemannian flowATLAS experimentOn-shell conditionParticle kinematicsDilepton resonance
Authors
Midori Kato, Kevin A. Urquía-Calderón, Inar Timiryasov, Oleg Ruchayskiy
Abstract
In this work we demonstrate that a single transformer-based generative model can capture Standard Model structure spanning five decades of invariant mass, from the sub-GeV regime to the TeV continuum, a range that no single Monte Carlo sample covers. To achieve this we design \textsc{ShellFlow}, a Riemannian conditional flow matching model that, given the recorded event composition, generates each particle on its on-shell manifold. Its only physics priors are the on-shell condition and the invariant-mass formula. The model is trained on $\sim 10^{9}$ real $pp$ collision events from the ATLAS Open Data 13~TeV release and told nothing else. From a single training run, the model learns to reproduce all of the following: intra-particle kinematics, the dilepton resonances ($J/ψ$, $Υ$, $Z$) at their PDG positions, the leptonic Weinberg angle, the $W$ and top-quark masses, and inter-particle correlations that enter no training objective. A substantial fraction of the Standard Model is thus learnable directly from recorded collision data.