Erlangen 2026 – scientific programme

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

T: Fachverband Teilchenphysik

T 15: Searches/BSM I

T 15.4: Talk

Monday, March 16, 2026, 17:00–17:15, KH 02.018

Anomaly detection for long-lived particles using CATHODE with CMS — •Jovin Drews, Louis Moureaux, Gregor Kasieczka, Karim El Morabit, Chitrakshee Yede, and Tore von Schwartz — University of Hamburg, Hamburg, Germany

In high-energy physics, numerous analyses search for phenomena beyond the Standard Model (BSM). Anomaly detection methods offer a model-independent way to probe such potential BSM signatures. We discuss CATHODE, an anomaly detection method combining density estimation and weak supervision, applied to signals from long-lived particles (LLP) embedded in multijet background from the CMS experiment. The work studies LLP scenarios with varying properties and different approaches to classification.

Keywords: CATHODE; Anomaly detection; Long-lived particles; CMS; Machine learning