aiqbench analysis
This research applies supervised and self-supervised neural embeddings to detect anomalies in simulated High-Luminosity LHC collision data, demonstrating that labeled training objectives generally outperform unsupervised ones for signal efficiency. The authors introduce a diagnostic pipeline using Gaussian Mixture Models to interpret flagged events within the Standard Model latent space, effectively recovering defining kinematics without relying on pre-defined observables. This work advances model-agnostic search strategies by bridging the gap between learned representations and physical interpretability essential for new-physics discovery.
Local Diagnostics of Learned Event Representations for Anomaly Detection at the HL-LHC
arXiv:2610.13437v1 Announce Type: new 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.
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.
Artificial Intelligence (cs.AI); High Energy Physics - Experiment (hep-ex); High Energy Physics - Phenomenology (hep-ph); Data Analysis, Statistics and Probability (physics.data-an)
arXiv:2610.13437 [cs.AI]
(or arXiv:2610.13437v1 [cs.AI] for this version)
https://doi.org/10.48550/arXiv.2610.13437 arXiv-issued DOI via DataCite (pending registration)
Full text under the publisher's open licence · 290 words
Related coverage
PAI: Preserving Amplitude Information in Representation-Based Time-Series Anomaly Detection
arXiv · Jun 9, 2026
AnomaMind: Agentic Time Series Anomaly Detection with Tool-Augmented Reasoning
arXiv · Jun 10, 2026
Localizing Global Discrepancies: Marginal Contributions and Contextual Anomaly Detection
arXiv · Aug 31, 2026
Vertical Advantage: Transforming Industries with Lakebase and Agentic AI
Databricks · Aug 27, 2026