Robustness of Neural Ratio and Posterior Estimators to Distributional Shifts for Population-Level Dark Matter Analysis in Strong Gravitational Lensing

Abstract

We investigate the robustness of Neural Ratio Estimators (NREs) and Neural Posterior Estimators (NPEs) to distributional shifts in the context of measuring the abundance of dark matter subhalos using strong gravitational lensing data. While these data-driven inference frameworks can be accurate on test data from the same distribution as the training sets, in real applications, it is expected that simulated training data and true observational data will differ in their distributions. We explore the behavior of a trained NRE and trained sequential NPEs to estimate the population-level parameters of dark matter subhalos from a large sample of images of strongly lensed galaxies with test data presenting distributional shifts within and beyond the bounds of the training distribution in the nuisance parameters (e.g., the background source morphology). While our results show that NREs and NPEs perform well when tested perfectly in distribution, they exhibit significant biases when confronted with slight deviations from the examples seen in the training distribution. This indicates the necessity for caution when applying NREs and NPEs to real astrophysical data, where high-dimensional underlying distributions are not perfectly known.

20 pages, 7 figures, 4 tables, accepted at ApJ, Volume 989, Issue 2, id.226, 15 pp

References in corpus (19)

• lenstronomy II: A gravitational lensing software ecosystem
• The shapes, orientation, and alignment of Galactic dark matter subhalos
• The Sloan Lens ACS Survey. XIII. Discovery of 40 New Galaxy-Scale Strong Lenses
• The BOSS Emission-Line Lens Survey. IV. : Smooth Lens Models for the BELLS GALLERY Sample
• Size bias and differential lensing of strongly lensed, dusty galaxies identified in wide-field surveys
• Mining for Dark Matter Substructure: Inferring subhalo population properties from strong lenses with machine learning
• AstroCLIP: A Cross-Modal Foundation Model for Galaxies
• IllustrisTNG in the HSC-SSP: image data release and the major role of mini mergers as drivers of asymmetry and star formation
• Hierarchical Inference With Bayesian Neural Networks: An Application to Strong Gravitational Lensing
• On the density profile of dark matter substructure in gravitational lens galaxies
• From Images to Dark Matter: End-To-End Inference of Substructure From Hundreds of Strong Gravitational Lenses
• The Completed SDSS-IV extended Baryon Oscillation Spectroscopic Survey: A Catalogue of Strong Galaxy-Galaxy Lens Candidates
• SICRET: Supernova Ia Cosmology with truncated marginal neural Ratio EsTimation
• Strong lensing selection effects
• One never walks alone: the effect of the perturber population on subhalo measurements in strong gravitational lenses
• Not All Subhaloes Are Created Equal: Modelling the Diversity of Subhalo Density Profiles in TNG50
• Subhalo effective density slope measurements from HST strong lensing data with neural likelihood-ratio estimation
• Inferring subhalo effective density slopes from strong lensing observations with neural likelihood-ratio estimation
• Tackling the Problem of Distributional Shifts: Correcting Misspecified, High-Dimensional Data-Driven Priors for Inverse Problems

Cited by in corpus (1)

• LSST Strong Lensing Systems Dark Matter Sensitivity Analysis with Neural Ratio Estimators