English

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

High Energy Physics - Phenomenology 2025-02-07 v4

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.

Keywords

Cite

@article{arxiv.2312.03067,
  title  = {Semi-visible jets, energy-based models, and self-supervision},
  author = {Luigi Favaro and Michael Krämer and Tanmoy Modak and Tilman Plehn and Jan Rüschkamp},
  journal= {arXiv preprint arXiv:2312.03067},
  year   = {2025}
}

Comments

18 pages, 6 figures, journal version

R2 v1 2026-06-28T13:42:09.243Z