Local Diagnostics of Learned Event Representations for Anomaly Detection at the HL-LHC

Abstract

Anomaly detection, the identification of collision events that deviate from Standard Model expectations, is one of the most promising model-agnostic strategies for new-physics searches at the Large Hadron Collider. Searching for rare events in learned low-dimensional representations removes the need to commit to hand-crafted observables a priori, but sacrifices the immediate physical interpretation those observables provide. For anomaly detection to enter a discovery pipeline, flagged events must therefore be diagnosed locally, both to distinguish potential new physics from mismodeling or detector effects and to identify observables for targeted follow-up searches. We evaluate supervised and self-supervised neural embedding strategies on COLLIDE-2V, a 750-million-event dataset simulated under realistic High-Luminosity LHC conditions, using the reconstruction error of an autoencoder trained in the learned latent space as the anomaly score. Across the configurations and benchmark signals considered, supervised objectives, which exploit simulation process labels during training, generally outperform their self-supervised counterparts and, in their best configurations, improve on an autoencoder trained directly on kinematic features, raising the signal efficiency at 10% background acceptance on all held-out processes. We further introduce a local diagnostic pipeline that fits a Gaussian Mixture Model to the SM latent space, profiles each component through high-level observables without process labels, localizes the flagged events, and ranks observables by their local discriminating power. The ranking recovers the defining kinematics of the held-out processes, while correlating the anomaly score with the same observables quantifies how closely the score tracks them, revealing cases where the most discriminating observable is not the one the score follows most closely.

Readable Is Not Steerable: Pre-Registered Tests of Physical Structure in Learned Fluid Representations
Manvendra Modgil
arXiv3 Nov 2026PreprintCC BY 4.0Full textFluid dynamicsarXiv
Conservative Orbital Resonance Networks as Local Environmental Diagnostics for Extreme Mass Ratio Inspirals
Mona Kumari, Anirudh Pradhan, K. Ghaderi +1
arXiv16 Oct 2026PreprintCC BY 4.0Full textGravitation & general relativityarXiv
Anomaly detection for multijet scenarios
Gregor Kasieczka, Sung Hak Lim, Louis Moureaux +3
arXiv14 Oct 2026PreprintCC BY 4.0Full textComputational physics High-energy & particle physicsarXiv
Domain Adaptation against Background Sculpting in Anomaly Detection at the LHC
Vincent Benne, Marie Hein, Michael Krämer +3
arXiv15 Sep 2026PreprintCC BY 4.0Full textHigh-energy & particle physicsarXiv
Model-independent searches and anomaly detection at the CMS experiment
Tamas Almos Vami, for the CMS Collaboration
arXiv22 Sep 2026PreprintCC BY 4.0Full textHigh-energy & particle physicsarXiv

Cite

BibTeX RIS

BibTeX

@misc{genovese2026local,
  title = {{Local Diagnostics of Learned Event Representations for Anomaly Detection at the HL-LHC}},
  author = {Donatella Genovese and Philip Ploner and Simone Scardapane and Stefano Giagu and Thea Klaeboe Aarrestad},
  howpublished = {arXiv},
  year = {2026},
  month = oct,
  doi = {10.48550/arxiv.2610.13437},
  eprint = {2610.13437},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2610.13437},
  note = {CC BY 4.0},
}

RIS

TY  - UNPB
TI  - Local Diagnostics of Learned Event Representations for Anomaly Detection at the HL-LHC
AU  - Donatella Genovese
AU  - Philip Ploner
AU  - Simone Scardapane
AU  - Stefano Giagu
AU  - Thea Klaeboe Aarrestad
PY  - 2026
DA  - 2026/10/09
PB  - arXiv
DO  - 10.48550/arxiv.2610.13437
UR  - https://arxiv.org/abs/2610.13437
AB  - Anomaly detection, the identification of collision events that deviate from Standard Model expectations, is one of the most promising model-agnostic strategies for new-physics searches at the Large Hadron Collider. Searching for rare events in learned low-dimensional representations removes the need to commit to hand-crafted observables a priori, but sacrifices the immediate physical interpretation those observables provide. For anomaly detection to enter a discovery pipeline, flagged events must therefore be diagnosed locally, both to distinguish potential new physics from mismodeling or detector effects and to identify observables for targeted follow-up searches. We evaluate supervised and self-supervised neural embedding strategies on COLLIDE-2V, a 750-million-event dataset simulated under realistic High-Luminosity LHC conditions, using the reconstruction error of an autoencoder trained in the learned latent space as the anomaly score. Across the configurations and benchmark signals considered, supervised objectives, which exploit simulation process labels during training, generally outperform their self-supervised counterparts and, in their best configurations, improve on an autoencoder trained directly on kinematic features, raising the signal efficiency at 10% background acceptance on all held-out processes. We further introduce a local diagnostic pipeline that fits a Gaussian Mixture Model to the SM latent space, profiles each component through high-level observables without process labels, localizes the flagged events, and ranks observables by their local discriminating power. The ranking recovers the defining kinematics of the held-out processes, while correlating the anomaly score with the same observables quantifies how closely the score tracks them, revealing cases where the most discriminating observable is not the one the score follows most closely.
ER  -

Abstract

Related

Cite