Hybrid quantum classical graph neural networks for particle track reconstruction

Abstract The Large Hadron Collider (LHC) at the European Organisation for Nuclear Research (CERN) will be upgraded to further increase instantaneous rate of particle collisions (luminosity) and become High Luminosity LHC (HL-LHC). This in luminosity significantly number particles interacting with detector. interaction a detector is referred as “hit”. HL-LHC yield many more hits, which pose combinatorial challenge by using reconstruction algorithms determine trajectories from those hits. work explores possibility converting novel graph neural network model, that can optimally take into account sparse nature tracking data their complex geometry, hybrid quantum-classical benefits variational quantum layers. We show this model perform similar classical approach. Also, we explore parametrized circuits (PQC) different expressibility entangling capacities, compare training performance order quantify expected benefits. These results used build future road map develop circuit-based networks.