SIMSHIFT: A Benchmark for Adapting Neural Surrogates to Distribution Shifts

Paul Setinek, Gianluca Galletti, Thomas Gross, Dominik Schnürer, Johannes Brandstetter, Werner Zellinger

Keywords: PDE, Neural Operators, Domain Adaptation, Distribution Shift

TL;DR: We introduce SIMSHIFT, a benchmark and dataset for domain adaptation of neural surrogates across four industrial simulation tasks with predefined distribution shifts.

Abstract: Neural surrogates for Partial Differential Equations (PDEs) often exhibit significant performance degradation when evaluated on unseen problem configurations, such as unseen material types or structural dimensions. Meanwhile, Domain Adaptation (DA) techniques have been extensively applied in vision and language processing to enable generalization from limited information about unseen configurations. In this work, we address this gap through two focused contributions. First, we introduce SIMSHIFT, a novel benchmark dataset and evaluation suite composed of four industrial simulation tasks: *hot rolling*, *sheet metal forming*, *electric motor design* and *heatsink design*. Second, we extend established domain adaptation methods to state of the art neural surrogates and systematically evaluate them. These approaches use parametric descriptions and ground-truth simulations from multiple source configurations, together with only parametric descriptions from target configurations. The goal is to accurately predict target simulations without access to ground-truth simulation data. Extensive experiments on SIMSHIFT highlight the challenges of out of distribution neural surrogate modeling, the potential of DA in simulation, and open problems for achieving robust neural surrogate modeling under distribution shifts in industrially relevant use cases.

Dataset URL: https://huggingface.co/datasets/simshift/SIMSHIFT_data

Primary Area: AL/ML Datasets & Benchmarks for physics (e.g. climate, health, life sciences, physics, social sciences)

Submission Number: 1437