Saved in:
Bibliographic Details
Main Authors: Visive, Ambre, Moskvitina, Polina, Nellist, Clara, de Austri, Roberto Ruiz, Caron, Sascha
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2509.26218
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912853187887104
author Visive, Ambre
Moskvitina, Polina
Nellist, Clara
de Austri, Roberto Ruiz
Caron, Sascha
author_facet Visive, Ambre
Moskvitina, Polina
Nellist, Clara
de Austri, Roberto Ruiz
Caron, Sascha
contents We propose a novel use of Large Language Models (LLMs) as unsupervised anomaly detectors in particle physics. Using lightweight LLM-like networks with encoder-based architectures trained to reconstruct background events via masked-token prediction, our method identifies anomalies through deviations in reconstruction performance, without prior knowledge of signal characteristics. Applied to searches for simultaneous four-top-quark production, this token-based approach shows competitive performance against established unsupervised methods and effectively captures subtle discrepancies in collider data, suggesting a promising direction for model-independent searches for new physics.
format Preprint
id arxiv_https___arxiv_org_abs_2509_26218
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Event Tokenization and Masked-Token Prediction for Anomaly Detection at the Large Hadron Collider
Visive, Ambre
Moskvitina, Polina
Nellist, Clara
de Austri, Roberto Ruiz
Caron, Sascha
High Energy Physics - Experiment
High Energy Physics - Phenomenology
We propose a novel use of Large Language Models (LLMs) as unsupervised anomaly detectors in particle physics. Using lightweight LLM-like networks with encoder-based architectures trained to reconstruct background events via masked-token prediction, our method identifies anomalies through deviations in reconstruction performance, without prior knowledge of signal characteristics. Applied to searches for simultaneous four-top-quark production, this token-based approach shows competitive performance against established unsupervised methods and effectively captures subtle discrepancies in collider data, suggesting a promising direction for model-independent searches for new physics.
title Event Tokenization and Masked-Token Prediction for Anomaly Detection at the Large Hadron Collider
topic High Energy Physics - Experiment
High Energy Physics - Phenomenology
url https://arxiv.org/abs/2509.26218