Does Unsupervised Domain Adaptation Improve the Robustness of Amortized Bayesian Inference? A Systematic Evaluation

TL;DR

This paper systematically evaluates whether unsupervised domain adaptation enhances the robustness of neural amortized Bayesian inference across various domain mismatches, revealing both benefits and limitations in different scenarios.

Contribution

It provides a comprehensive analysis of UDA's effectiveness in improving ABI robustness, highlighting conditions where it succeeds or fails, which was previously underexplored.

Findings

UDA can mitigate effects of unmodeled noise and phenomena

Alignment may fail under prior misspecification

Careful consideration needed for different mismatch types

Abstract

Neural networks are fragile when confronted with data that significantly deviates from their training distribution. This is true in particular for simulation-based inference methods, such as neural amortized Bayesian inference (ABI), where models trained on simulated data are deployed on noisy real-world observations. Recent robust approaches employ unsupervised domain adaptation (UDA) to match the embedding spaces of simulated and observed data. However, the lack of comprehensive evaluations across different domain mismatches raises concerns about the reliability in high-stakes applications. We address this gap by systematically testing UDA approaches across a wide range of misspecification scenarios in silico and practice. We demonstrate that aligning summary spaces between domains effectively mitigates the impact of unmodeled phenomena or noise. However, the same alignment mechanism can lead to failures under prior misspecifications - a critical finding with practical consequences. Our results underscore the need for careful consideration of misspecification types when using UDA to increase the robustness of ABI.