“…Recently, there has been significant interest in the use of supervised machine learning (ML) to perform the reconstruction in order to improve the physics reach of the experiments as well as reduce the computational requirements. ML-based reconstruction approaches using graph neural networks (GNNs) [14,15,16,17,18] have been proposed for various tasks in particle physics [19,20,21,22,23,24,25,26,27,28,29,30,31,32,33], including PF reconstruction [34,35,36,37]. In particular in Ref.…”