Variational Autoencoders for Anomalous Jet Tagging

TL;DR

This paper explores the use of Variational Autoencoders for detecting anomalous jets at the LHC, introducing the OE-VAE to improve anomaly detection sensitivity and mass decorrelation.

Contribution

It introduces the Outlier Exposed VAE (OE-VAE), a novel approach that incorporates outlier samples during training to enhance anomaly detection and jet mass decorrelation.

Findings

OE-VAE improves anomaly detection sensitivity.

Mass decorrelation achieved without sacrificing detection performance.

Naive mass-decorrelated VAEs fail to maintain proper detection performance.

Abstract

We present a detailed study on Variational Autoencoders (VAEs) for anomalous jet tagging at the Large Hadron Collider. By taking in low-level jet constituents' information, and training with background QCD jets in an unsupervised manner, the VAE is able to encode important information for reconstructing jets, while learning an expressive posterior distribution in the latent space. When using the VAE as an anomaly detector, we present different approaches to detect anomalies: directly comparing in the input space or, instead, working in the latent space. In order to facilitate general search approaches such as bump-hunt, mass-decorrelated VAEs based on distance correlation regularization are also studied. We find that the naive mass-decorrelated VAEs fail at maintaining proper detection performance, by assigning higher probabilities to some anomalous samples. To build a performant mass-decorrelated anomalous jet tagger, we propose the Outlier Exposed VAE (OE-VAE), for which some outlier samples are introduced in the training process to guide the learned information. OE-VAEs are employed to achieve two goals at the same time: increasing sensitivity of outlier detection and decorrelating jet mass from the anomaly score. We succeed in reaching excellent results from both aspects. Code implementation of this work can be found at https://github.com/taolicheng/VAE-Jet