%0 Journal Article
%T Fast b-tagging at the high-level trigger of the ATLAS experiment in LHC Run 3
%A ATLAS Collaboration (Aad, G. et al
%A Akiot, A.
%A Amos, K. R.
%A Aparisi Pozo, J. A.
%A Bailey, A. J.
%A Bouchhar, N.
%A Cabrera Urban, S.
%A Cantero, J.
%A Cardillo, F.
%A Castillo Gimenez, V.
%A Chitishvili, M.
%A Costa, M. J.
%A Didenko
%A Escobar, C.
%A Fiorini, L.
%A Fullana Torregrosa, E.
%A Fuster, J.
%A Garcia, C.
%A Garcia Navarro, J. E.
%A Gomez Delegido, A. J.
%A Gonzalez de la Hoz, S.
%A Gonzalvo Rodriguez, G. R.
%A Guerrero Rojas, J. G. R.
%A Lacasta, C.
%A Marti-Garcia, S.
%A Martinez Agullo, P.
%A Miralles Lopez, M.
%A Mitsou, V. A.
%A Monsonis Romero, L.
%A Moreno Llacer, M.
%A Munoz Perez, D.
%A Navarro-Gonzalez, J.
%A Poveda, J.
%A Prades Ibañez, A.
%A Rubio Jimenez, A.
%A Ruiz-Martinez, A.
%A Sabatini, P.
%A Salt, J.
%A Sanchez Sebastian, V.
%A Sayago Galvan, I.
%A Senthilkumar, V.
%A Soldevila, U.
%A Sanchez, J.
%A Torro Pastor, E.
%A Valero, A.
%A Valiente Moreno, E.
%A Valls Ferrer, J. A.
%A Varriale, L.
%A Villaplana Perez, M.
%A Vos, M.
%J Journal of Instrumentation
%D 2023
%V 18
%N 11
%I IOP Publishing Ltd
%@ 1748-0221
%G English
%F ATLASCollaborationAad_etal2023
%O WOS:001123791900004
%O exported from refbase (https://references.ific.uv.es/refbase/show.php?record=5972), last updated on Fri, 15 Mar 2024 08:52:50 +0000
%X The ATLAS experiment relies on real-time hadronic jet reconstruction and b-tagging to record fully hadronic events containing b-jets. These algorithms require track reconstruction, which is computationally expensive and could overwhelm the high-level-trigger farm, even at the reduced event rate that passes the ATLAS first stage hardware-based trigger. In LHC Run 3, ATLAS has mitigated these computational demands by introducing a fast neural-network-based b-tagger, which acts as a low-precision filter using input from hadronic jets and tracks. It runs after a hardware trigger and before the remaining high-level-trigger reconstruction. This design relies on the negligible cost of neural-network inference as compared to track reconstruction, and the cost reduction from limiting tracking to specific regions of the detector. In the case of Standard Model HH -> b (b) over barb (b) over bar, a key signature relying on b-jet triggers, the filter lowers the input rate to the remaining high-level trigger by a factor of five at the small cost of reducing the overall signal efficiency by roughly 2%.
%K Trigger algorithms
%K Trigger concepts and systems (hardware and software)
%R 10.1088/1748-0221/18/11/P11006
%U https://arxiv.org/abs/2306.09738
%U https://doi.org/10.1088/1748-0221/18/11/P11006
%P P11006 - 38pp