Semi-visible jets, energy-based models, and self-supervision

SciPost Physics, vol. 18

Abstract

We present DarkCLR, a novel framework for detecting semi-visible jets at the LHC. DarkCLR uses a self-supervised contrastive-learning approach to create observables that are approximately invariant under relevant transformations. We use background-enhanced data to create a sensitive representation and evaluate the representations using a CLR-inspired anomaly score and a normalized autoencoder as density estimators. Our results show a remarkable sensitivity for a wide range of semi-visible jets and are more robust than a supervised classifier trained on a specific signal.

Authors 5

1. Luigi Favaro

   Heidelberg University

   Affiliation as printed

   Heidelberg University

2. Michael Krämer Aachen

   RWTH Aachen University

   Affiliation as printed

   RWTH Aachen University

3. Tanmoy Modak

   Heidelberg University

   Affiliation as printed

   Heidelberg University

4. Tilman Plehn

   Heidelberg University

   Affiliation as printed

   Heidelberg University

5. Jan Rüschkamp

   Heidelberg University

   Affiliation as printed

   Heidelberg University