Karlsruhe 2024 – scientific programme
Parts | Days | Selection | Search | Updates | Downloads | Help
T: Fachverband Teilchenphysik
T 94: Trigger+DAQ 3
T 94.5: Talk
Thursday, March 7, 2024, 17:00–17:15, Geb. 30.23: 3/1
Elevating LHCb’s beauty selection for Run 3: A neural network approach — Johannes Albrecht1, Gregory Max Ciezarek2, Blaise Delaney3, Niklas Nolte4, and •Nicole Schulte1 — 1TU Dortmund University, Dortmund, Germany — 2CERN, Geneva, Switzerland — 3Massachusetts Institute of Technology, Cambridge, USA — 4META AI (FAIR)
The performance of LHCb’s beauty physics program relies significantly on b-hadron triggers, specifically topological triggers. These triggers are designed for the comprehensive identification of b-hadron candidates, leveraging the distinct decay topology of beauty particles and their anticipated kinematic properties. Constituting the predominant component on the trigger selection output, topological triggers play a crucial role in the success of numerous physics analyses within LHCb.
In this contribution, we present the Run 3 implementation of the topological trigger, seamlessly integrating Lipschitz monotonic neural networks. This architecture ensures resilience in the face of varying detector conditions and enhances sensitivity to long-lived candidates. This framework can potentially open avenues for the discovery of new physics at LHCb. The primary focus is on synergizing a comprehensive physics selection with state-of-the-art machine learning approaches, all within the constraints of available computational resources.
Keywords: LHCb; Trigger; Machine Learning; Neural Networks; Real-Time