A deep neural network-based tagger to search for new long-lived particle states decaying to jets

The development of a tagging algorithm to identify jets that are significantly displaced from the luminous regions of LHC proton-proton (pp) collisions is presented. Displaced jets can arise from the decay of a long-lived particle (LLP), which are predicted by several theoretical extensions to the standard model. The tagger is a multiclass classifier based on a deep neural network, which is parameterized according to the proper decay length $\text{c}\tau_0$ of the LLP. A novel scheme is defined to reliably label jets from LLP decays for supervised learning. Samples of both simulated events and pp collision data are used to train the neural network. Domain adaptation by backward propagation is performed to improve the simulation modelling of the jet class probability distributions observed in pp collision data. The tagger is applied in a search for long-lived gluinos, a manifestation of split supersymmetric models. The tagger provides a rejection factor of $10\,000$ for jets from standard model processes while maintaining an LLP jet tagging efficiency of 30${-}$80$\%$ for split supersymmetric models with $1\,\text{mm} \leq \text{c}\tau_0 \leq 10\,\text{m}$. The expected coverage of the split supersymmetric model parameter space is presented.