Learning the Standard Model Manifold: Bayesian Latent Diffusion for Collider Anomaly Detection

TL;DR

This paper introduces a physics-informed Bayesian latent diffusion model for collider data anomaly detection, improving stability and generalization by integrating physics constraints and uncertainty estimation.

Contribution

It presents a novel diffusion-based framework that combines probabilistic encoding with physics-motivated regularization for collider anomaly detection.

Findings

Diffusion process enhances training stability and generalization.

Physics constraints improve anomaly detection reliability.

Model achieves competitive ROC and discovery metrics on simulated data.

Abstract

We propose a physics-informed anomaly detection framework for collider data based on a Bayesian latent diffusion model. Our method combines a probabilistic encoder with diffusion dynamics in the latent space, allowing for stable and flexible density estimation while explicitly enforcing physics constraints, such as mass decorrelation and regularization of latent correlations. We train and test the model on simulated LHC jet data and evaluate its performance using seed-averaged ROC curves together with discovery-oriented metrics. Through a series of ablation studies, we show that the diffusion process, Bayesian regularization, and physics-motivated loss terms each contribute in a complementary way: they help stabilize training and improve generalization, even when the gains in peak performance are moderate. Overall, our results emphasize the importance of incorporating both uncertainty estimates and physics consistency when building reliable anomaly detection methods for new Physics searches in high-energy physics.