Parts | Days | Selection | Search | Updates | Downloads | Help

T: Fachverband Teilchenphysik

T 91: Methods in Particle Physics V

T 91.4: Talk

Friday, March 20, 2026, 09:45–10:00, KH 00.020

Background suppression of neutrino inelastic interactions in SHiP with a GNN using the Surrounding Background Tagger — •Darja Lund for the SHiP-SBT collaboration — Humboldt Universität zu Berlin, Berlin, Germany

The SHiP (Search for Hidden Particles) fixed target experiment will be realised in a dedicated beam-dump facility in the ECN3 cavern in CERN's north area. It will take advantage of the 400 GeV SPS proton beam to search for feebly interacting long-lived particles at the MeV-GeV scale over many orders of magnitude in coupling. The search for decay signatures will be facilitated by the full reconstruction and particle identification of Standard model final states using the detector setup. Located downstream of the hadron stopper and the magnetic muon shield is the 50m long decay volume, encased by the SBT (Surrounding Background Tagger). A major focus of the experiment, and a key functionality of the SBT, is the detection and suppression of muon and neutrino inelastic scattering background inside the 50m long decay volume. For this, a GNN (Graph Neural Network) to identify background is in constant development. This talk will give an overview of the neutrino background suppression using a GNN-based veto.

Keywords: Background Suppression; SHiP; Machine Learning