Additional QCD Background Events for LHCO2020 R&D (signal region only)

These are an additional ~610k QCD dijet background events, in the "signal region" region only (meaning, with mJJ in the range 3.3-3.7 TeV), produced using the same simulation settings as those of the LHCO2020 R&D Dataset. The features are as described at that link:

'pxj1', 'pyj1', 'pzj1', 'mj1', 'tau1j1', 'tau2j1', 'tau3j1', 'pxj2', 'pyj2', 'pzj2', 'mj2', 'tau1j2', 'tau2j2', 'tau3j2'

They were used in CATHODE for training the idealized anomaly detector and supervised classifier, as well as for final evaluation of all the anomaly detection methods. 

Update September 22, 2023

An additional file was uploaded that contains the momentum 3-vectors of all the particles in each event, in addition to the 14 high level features listed above. As in the original LHCO R&D dataset, up to 700 particles are included, with zero padding for events with fewer than 700 particles. So the shape of the dataset is 

(612858, 2114)

with the first 2100 entries corresponding to the 3-momenta, in the order px0,py0,pz0,px1,py1,pz1,...,px699,py699,pz699; and the last 14 entries corresponding to the high level features listed above.