Multi-Prong Dataset

Simulated events for the classification of jets with multiple hard prongs. Total of ~100K events, split between seven classes. Description 13 TeV events simulated with MadGraph5 Showering and hadronization simulated with Pythia8 Detector response simulation with Delphes using an ATLAS-like card Jets are clustered using the anti-kT algorithm with radius parameter R=1.2 using FastJet The pT of the jets lies in the range [1000, 1200] GeV The mass of the jets lies in the range [300, 700] GeV The three-momenta of the leading 230 constituents are stored in the 'Constituents/threeM' dataset Jets with fewer than 230 constituents are zero-padded Truth labels are stored in the 'target' dataset Labels 0/1/2/3/4/5/6 correspond to jets with N=1/2/3/4b/6/8/4q sub-jets, respectively See arXiv:2202.00723 for more details