24–28 Aug 2026
Kirchhoff Institute for Physics (KIP)
Europe/Berlin timezone

Towards a Statistical Interpretation of the Normalized Autoencoder

24 Aug 2026, 17:00
8m
3.404

3.404

Patterns & Anomalies 🔀 Patterns & Anomalies

Speaker

Jonathan Ostertag-Henning (ITP Heidelberg)

Description

Unsupervised anomaly detection with autoencoders is a promising data-driven and model-agnostic approach for new physics searches at the LHC. However, current anomaly scores assigned by neural networks suffer from a lack of statistical interpretability. The normalized autoencoder (NAE) combines a standard bottleneck architecture with a well-defined probabilistic description. We show that the NAE ties its anomaly score to a learned likelihood via an energy-based training objective, and introduce a Bayesian version of the NAE (BNAE) that additionally provides machine-learned uncertainty estimates. We validate both on a toy model and demonstrate competitive and symmetric anomaly-tagging performance on top-versus-QCD jet tagging.

Authors

Ranit Das (Heidelberg University) Jonathan Ostertag-Henning (ITP Heidelberg) Tilman Plehn (Institut für Theoretische Physik, Universität Heidelberg; Interdisciplinary Center for Scientific Computing (IWR), Universität Heidelberg) Lorenz Vogel (Institute for Theoretical Physics (ITP), Heidelberg University, Germany)

Presentation materials