Göttingen 2025 – wissenschaftliches Programm

Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

T: Fachverband Teilchenphysik

T 77: Data, AI, Computing, Electronics VIII (Fast ML, Triggers)

T 77.6: Vortrag

Donnerstag, 3. April 2025, 17:30–17:45, VG 2.102

Development of machine-learning based topological triggers for the CMS Level-1 trigger — •Karla Kleinbölting, Lukas Ebeling, Johannes Haller, Finn Jonathan Labe, Balduin Letzer, Artur Lobanov, Lara Markus, and Matthias Schröder — Institut für Experimentalphyisk, Universität Hamburg

At the CMS experiment, the Level-1 (L1) trigger system is pivotal in thereal-timeselectionofphysicseventsofinterest. Thistalkhighlights recent advancements in enhancing the L1 trigger performance through the integration of machine learning (ML) techniques. Using di-Higgs production as a benchmark process, the proposed ML-based trigger leverages full event topologies instead of individual object-based trig- gers. This approach allows the trigger system to identify and retain events in previously inaccessible low p_T regions while maintaining ac- ceptable rates. The ML algorithms can be seamlessly integrated into the FPGA-based electronics of the trigger system using frameworks such as hls4ml.

Keywords: Level-1 trigger; Machine Learning; di-Higgs; Topological algorithms