Better latent spaces for better autoencoders

Dillon, Barry M.; Plehn, Tilman; Sauer, Christof; Sorrenson, Peter

doi:10.21468/SciPostPhys.11.3.061

SciPost Physics

Better latent spaces for better autoencoders

Barry M. Dillon, Tilman Plehn, Christof Sauer, Peter Sorrenson

SciPost Phys. 11, 061 (2021) · published 17 September 2021

doi: 10.21468/SciPostPhys.11.3.061
pdf
Submissions/Reports

Abstract

Autoencoders as tools behind anomaly searches at the LHC have the structural problem that they only work in one direction, extracting jets with higher complexity but not the other way around. To address this, we derive classifiers from the latent space of (variational) autoencoders, specifically in Gaussian mixture and Dirichlet latent spaces. In particular, the Dirichlet setup solves the problem and improves both the performance and the interpretability of the networks.

TY  - JOUR
PB  - SciPost Foundation
DO  - 10.21468/SciPostPhys.11.3.061
TI  - Better latent spaces for better autoencoders
PY  - 2021/09/17
UR  - https://scipost.org/SciPostPhys.11.3.061
JF  - SciPost Physics
JA  - SciPost Phys.
VL  - 11
IS  - 3
SP  - 061
A1  - Dillon, Barry M.
AU  - Plehn, Tilman
AU  - Sauer, Christof
AU  - Sorrenson, Peter
AB  - Autoencoders as tools behind anomaly searches at the LHC have the structural problem that they only work in one direction, extracting jets with higher complexity but not the other way around. To address this, we derive classifiers from the latent space of (variational) autoencoders, specifically in Gaussian mixture and Dirichlet latent spaces. In particular, the Dirichlet setup solves the problem and improves both the performance and the interpretability of the networks.
ER  -

@Article{10.21468/SciPostPhys.11.3.061,
	title={{Better latent spaces for better autoencoders}},
	author={Barry M. Dillon and Tilman Plehn and Christof Sauer and Peter Sorrenson},
	journal={SciPost Phys.},
	volume={11},
	pages={061},
	year={2021},
	publisher={SciPost},
	doi={10.21468/SciPostPhys.11.3.061},
	url={https://scipost.org/10.21468/SciPostPhys.11.3.061},
}

Cited by 31

Kasieczka et al., Anomaly detection under coordinate transformations
Phys. Rev. D 107, 015009 (2023) [Crossref]
Aarrestad et al., The Dark Machines Anomaly Score Challenge: Benchmark Data and Model Independent Event Classification for the Large Hadron Collider
SciPost Phys. 12, 043 (2022) [Crossref]
Aguilar-Saavedra, Anomaly detection from mass unspecific jet tagging
Eur. Phys. J. C 82, 130 (2022) [Crossref]
Ostdiek, Deep Set Auto Encoders for Anomaly Detection in Particle Physics
SciPost Phys. 12, 045 (2022) [Crossref]
Barron et al., Unsupervised hadronic SUEP at the LHC
J. High Energ. Phys. 2021, 129 (2021) [Crossref]
Karagiorgi et al., Machine learning in the search for new fundamental physics
Nat Rev Phys 4, 399 (2022) [Crossref]
Cheng et al., Invariant representation driven neural classifier for anti-QCD jet tagging
J. High Energ. Phys. 2022, 152 (2022) [Crossref]
Dillon et al., A normalized autoencoder for LHC triggers
SciPost Phys. Core 6, 074 (2023) [Crossref]
Fraser et al., Challenges for unsupervised anomaly detection in particle physics
J. High Energ. Phys. 2022, 66 (2022) [Crossref]
Sabry et al., Image Retrieval Using Convolutional Autoencoder, InfoGAN, and Vision Transformer Unsupervised Models
IEEE Access 11, 20445 (2023) [Crossref]
Krzyzanska et al., Simulation-based anomaly detection for multileptons at the LHC
J. High Energ. Phys. 2023, 61 (2023) [Crossref]
Kamal Afify et al.,
, 1 (2023) [Crossref]
Hallin et al., Classifying anomalies through outer density estimation
Phys. Rev. D 106, 055006 (2022) [Crossref]
Dillon et al., Symmetries, safety, and self-supervision
SciPost Phys. 12, 188 (2022) [Crossref]
Gonski et al., High-dimensional anomaly detection with radiative return in e+e− collisions
J. High Energ. Phys. 2022, 156 (2022) [Crossref]
Verheyen, Event Generation and Density Estimation with Surjective Normalizing Flows
SciPost Phys. 13, 047 (2022) [Crossref]
Aguilar-Saavedra, Taming modeling uncertainties with mass unspecific supervised tagging
Eur. Phys. J. C 82, 270 (2022) [Crossref]
Bradshaw et al., Creating simple, interpretable anomaly detectors for new physics in jet substructure
Phys. Rev. D 106, 035014 (2022) [Crossref]
Belis et al., Machine learning for anomaly detection in particle physics
Reviews in Physics 12, 100091 100091 (2024) [Crossref]
Bai et al., Non-resonant anomaly detection with background extrapolation
J. High Energ. Phys. 2024, 59 (2024) [Crossref]
Bortolato et al., Bump hunting in latent space
Phys. Rev. D 105, 115009 (2022) [Crossref]
Alvarez et al., Exploring unsupervised top tagging using Bayesian inference
SciPost Phys. Core 6, 046 (2023) [Crossref]
Buss et al., What's anomalous in LHC jets?
SciPost Phys. 15, 168 (2023) [Crossref]
Mikuni et al., Online-compatible unsupervised nonresonant anomaly detection
Phys. Rev. D 105, 055006 (2022) [Crossref]
Ngairangbam et al., Anomaly detection in high-energy physics using a quantum autoencoder
Phys. Rev. D 105, 095004 (2022) [Crossref]
Jiang et al., Detecting new physics as novelty — Complementarity matters
J. High Energ. Phys. 2022, 85 (2022) [Crossref]
Bonilla et al., Jets and Jet Substructure at Future Colliders
Front. Phys. 10, 897719 (2022) [Crossref]
Canelli et al., Autoencoders for semivisible jet detection
J. High Energ. Phys. 2022, 74 (2022) [Crossref]
Alvarez et al., Bayesian probabilistic modeling for four-top production at the LHC
Phys. Rev. D 105, 092001 (2022) [Crossref]
Atkinson et al., IRC-Safe Graph Autoencoder for Unsupervised Anomaly Detection
Front. Artif. Intell. 5, 943135 (2022) [Crossref]
Wittmer et al., An autoencoder compression approach for accelerating large-scale inverse problems
Inverse Problems 39, 115009 (2023) [Crossref]

Authors / Affiliation: mappings to Contributors and Organizations

See all Organizations.

¹ Barry M. Dillon,
¹ Tilman Plehn,
¹ Christof Sauer,
¹ Peter Sorrenson

¹ Ruprecht-Karls-Universität Heidelberg / Heidelberg University

Funders for the research work leading to this publication