Saved in:
Bibliographic Details
Main Authors: Zimmel, Anna, Setinek, Paul, Galletti, Gianluca, Brandstetter, Johannes, Zellinger, Werner
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2602.15820
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908838206111744
author Zimmel, Anna
Setinek, Paul
Galletti, Gianluca
Brandstetter, Johannes
Zellinger, Werner
author_facet Zimmel, Anna
Setinek, Paul
Galletti, Gianluca
Brandstetter, Johannes
Zellinger, Werner
contents Machine learning surrogates are increasingly used in engineering to accelerate costly simulations, yet distribution shifts between training and deployment often cause severe performance degradation (e.g., unseen geometries or configurations). Test-Time Adaptation (TTA) can mitigate such shifts, but existing methods are largely developed for lower-dimensional classification with structured outputs and visually aligned input-output relationships, making them unstable for the high-dimensional, unstructured and regression problems common in simulation. We address this challenge by proposing a TTA framework based on storing maximally informative (D-optimal) statistics, which jointly enables stable adaptation and principled parameter selection at test time. When applied to pretrained simulation surrogates, our method yields up to 7% out-of-distribution improvements at negligible computational cost. To the best of our knowledge, this is the first systematic demonstration of effective TTA for high-dimensional simulation regression and generative design optimization, validated on the SIMSHIFT and EngiBench benchmarks.
format Preprint
id arxiv_https___arxiv_org_abs_2602_15820
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Stabilizing Test-Time Adaptation of High-Dimensional Simulation Surrogates via D-Optimal Statistics
Zimmel, Anna
Setinek, Paul
Galletti, Gianluca
Brandstetter, Johannes
Zellinger, Werner
Machine Learning
Machine learning surrogates are increasingly used in engineering to accelerate costly simulations, yet distribution shifts between training and deployment often cause severe performance degradation (e.g., unseen geometries or configurations). Test-Time Adaptation (TTA) can mitigate such shifts, but existing methods are largely developed for lower-dimensional classification with structured outputs and visually aligned input-output relationships, making them unstable for the high-dimensional, unstructured and regression problems common in simulation. We address this challenge by proposing a TTA framework based on storing maximally informative (D-optimal) statistics, which jointly enables stable adaptation and principled parameter selection at test time. When applied to pretrained simulation surrogates, our method yields up to 7% out-of-distribution improvements at negligible computational cost. To the best of our knowledge, this is the first systematic demonstration of effective TTA for high-dimensional simulation regression and generative design optimization, validated on the SIMSHIFT and EngiBench benchmarks.
title Stabilizing Test-Time Adaptation of High-Dimensional Simulation Surrogates via D-Optimal Statistics
topic Machine Learning
url https://arxiv.org/abs/2602.15820