Graph Neural Network for Object Reconstruction in Liquid Argon Time Projection Chambers

This paper presents a graph neural network (GNN) technique for low-level reconstruction of neutrino interactions in Liquid Argon Time Projection Chamber (LArTPC). GNNs are still relatively novel technique, and have shown great promise similar tasks the LHC. In this paper, multihead attention message passing is used to classify relationship between detector hits by labelling edges, determining whether were produced same underlying particle, if so, particle type. The trained model 84% accurate overall, performs best on EM shower muon track classes. model's strengths weaknesses discussed, plans developing further summarised.