Uncovering Hidden Systematics in Neural Network Models for High Energy Physics

TL;DR

This paper investigates the hidden sensitivities of neural networks in high-energy physics, revealing their vulnerability to subtle, realistic input perturbations and proposing a framework to better estimate their systematic uncertainties.

Contribution

It introduces a novel framework to detect and quantify hidden sensitivities of neural networks to realistic input variations in high-energy physics analyses.

Findings

Neural networks can be systematically fooled within experimental uncertainties.

Subtle input perturbations can cause significant changes in network outputs.

The proposed framework helps evaluate and control systematic uncertainties.

Abstract

Neural networks (NNs) are inherently multidimensional classifiers that learn complex, non-linear relationships among input observables. While their flexibility enables unprecedented performance in high-energy physics (HEP) analyses, it also makes them sensitive to small variations in their inputs. Consequently, the propagation and estimation of systematic uncertainties in NN-based models remain an open challenge. There are indications that uncertainties derived in control regions or from nominal variations of input features can underestimate the true model uncertainty, potentially leaving biases unaccounted for. Inspired by insights from adversarial-attack studies in machine learning, we explore how subtle perturbations, fully consistent with the experimental uncertainties on the input observables, can lead to substantial changes in NN outputs, while keeping the one-dimensional and correlated input distributions nearly unchanged. Using a set of representative HEP tasks, including event classification and object identification, and testing across a variety of network architectures, we demonstrate that networks can be systematically "fooled" at significant rates within the allowed uncertainty envelopes. Building on this observation, we introduce a quantitative framework to probe and measure the hidden sensitivity of neural networks to realistic experimental variations, providing a practical path to evaluate and control their systematic uncertainty in physics analyses.