Towards replacing detector simulation with heterogeneous GNNs in flavour physics analyses

Abstract

Driven by the increasing volume of recorded data, the demand for simulation from experiments based at the Large Hadron Collider will rise sharply in the coming years. Addressing this demand solely with existing computationally intensive workflows is not feasible. This paper introduces a new fast simulation tool designed to address this demand at the LHCb experiment. This tool emulates the detector response to arbitrary multibody decay topologies at LHCb. Rather than memorising specific decay channels, the model learns generalisable patterns within the response, allowing it to interpolate to channels not present in the training data. Novel heterogeneous graph neural network architectures are employed that are designed to embed the physical characteristics of the task directly into the network structure. We demonstrate the performance of the tool across a range of decay topologies, showing the networks can correctly model the relationships between complex variables. The architectures and methods presented are generic and could readily be adapted to emulate workflows at other simulation-intensive particle physics experiments.

References in corpus (30)

• PYTHIA 6.4 Physics and Manual
• DELPHES 3, A modular framework for fast simulation of a generic collider experiment
• LHCb Detector Performance
• PHOTOS Monte Carlo: a precision tool for QED corrections in Z and W decays
• Theoretical predictions for charm and bottom production at the LHC
• Test of lepton universality in beauty-quark decays
• A Roadmap for HEP Software and Computing R&D for the 2020s
• Branching fraction measurements of the rare $B_s^0\to \varphi {\mu }^{+}{\mu }^{-}$ and $B_s^0\to f_2^{\prime }(1525){\mu }^{+}{\mu }^{-}$ decays
• Test of lepton universality in $b\to s{\ell }^{+}{\ell }^{-}$ decays
• Design and performance of the LHCb trigger and full real-time reconstruction in Run 2 of the LHC
• AtlFast3: the next generation of fast simulation in ATLAS
• RapidSim: an application for the fast simulation of heavy-quark hadron decays
• ReDecay: A novel approach to speed up the simulation at LHCb
• Measurement of the ratios of branching fractions $\mathcal{R}(D^{\ast })$ and $\mathcal{R}(D^0)$
• Test of lepton flavour universality using $B^0\to D^{\ast -}{\tau }^{+}{\nu }_{\tau }$ decays with hadronic $\tau$ channels
• Deep generative models for fast photon shower simulation in ATLAS
• Fast Data-Driven Simulation of Cherenkov Detectors Using Generative Adversarial Networks
• Measurement of the branching fraction ratios $R(D^{+})$ and $R(D^{\ast +})$ using muonic $\tau$ decays
• Heterogeneous Graph Neural Network for Identifying Hadronically Decayed Tau Leptons at the High Luminosity LHC
• A parametrized Kalman filter for fast track fitting at LHCb
• Measurement of the branching fraction ratio $R_K$ at large dilepton invariant mass
• Observation of the doubly-charmed-baryon decay ${\Xi }_{cc}^{++}\to {\Xi }_c^0{\pi }^{+}{\pi }^{+}$
• Reweighting simulated events using machine-learning techniques in the CMS experiment
• GAN with an Auxiliary Regressor for the Fast Simulation of the Electromagnetic Calorimeter Response
• First observation of the ${\Lambda }_b^0\to D^{+}{D}^{-}\Lambda$ decay
• Search for the lepton-flavor violating decay $B_s^0\to \varphi {\mu }^{\pm }{\tau }^{\mp }$
• Scalable Multi-Task Learning for Particle Collision Event Reconstruction with Heterogeneous Graph Neural Networks
• Measurement of $CP$ violation observables in $D^{+}\to K^{-}{K}^{+}{\pi }^{+}$ decays
• Measurement of the CKM angle $\gamma$ in $B^{\pm }\to D{K}^{\ast }(892{)}^{\pm }$ decays
• Study of $CP$ violation in $B_{(s)}^0\to D{K}^{\ast }(892{)}^0$ decays with $D\to K\pi (\pi \pi )$, $\pi \pi (\pi \pi )$, and $KK$ final states