Physics-informed fine-tuning of foundation models for partial differential equations

📅 2026-03-16
📈 Citations: 0
Influential: 0
📄 PDF

career value

233K/year
🤖 AI Summary
This work addresses the challenge of efficiently adapting pre-trained foundation models for partial differential equations (PDEs) under data scarcity and distribution shift. We propose a physics-informed fine-tuning framework that incorporates physical constraints—such as PDE residuals and boundary conditions—directly into the fine-tuning objective, enabling data-efficient adaptation without requiring ground-truth solutions. To our knowledge, this is the first systematic demonstration of the effectiveness of physics-informed fine-tuning for transferring PDE foundation models. By integrating a hybrid fine-tuning strategy, our approach significantly enhances out-of-distribution generalization. Experiments show that, even in the absence of ground-truth solutions, our method achieves accuracy comparable to purely data-driven approaches on unseen PDE tasks and consistently outperforms them when only limited data are available.

Technology Category

Application Category

📝 Abstract
Foundation models for partial differential equations (PDEs) have emerged as powerful surrogates pre-trained on diverse physical systems, but adapting them to new downstream tasks remains challenging due to limited task-specific data and distribution shifts. While fine-tuning has proven transformative in natural language processing, best practices for adapting PDE foundation models remain underexplored. Although physics-informed training has successfully trained accurate solvers across a wide range of PDE problems, its potential for fine-tuning data-based foundation models has not been systematically studied. In this work, we introduce a physics-informed fine-tuning framework that adapts pre-trained PDE foundation models by incorporating physical constraints (PDE residuals and boundary conditions) directly into the fine-tuning objective. This enables effective adaptation in data-scarce regimes while promoting physical consistency. We evaluate our method on a downstream task composed of an unseen PDE class and compare it with data-driven finetuning counterparts. Our results demonstrate that physics-informed fine-tuning achieves competitive accuracy without requiring PDE solutions for training. Furthermore, a hybrid fine-tuning strategy yields superior generalization to out-of-distribution scenarios when only minimal training data is available. These findings establish physics-informed fine-tuning as a scalable and data-efficient paradigm, providing a physically interpretable pathway for adapting foundation models in scientific machine learning.
Problem

Research questions and friction points this paper is trying to address.

physics-informed fine-tuning
foundation models
partial differential equations
data scarcity
distribution shift
Innovation

Methods, ideas, or system contributions that make the work stand out.

physics-informed fine-tuning
foundation models
partial differential equations
data-efficient adaptation
physical consistency
🔎 Similar Papers
V
Vlad Medvedev
Department Modeling and Artificial Intelligence, Fraunhofer Institute for Integrated Systems and Device Technology IISB, Schottkystrasse 10, 91058 Erlangen, Germany
L
Leon Armbruster
Department Modeling and Artificial Intelligence, Fraunhofer Institute for Integrated Systems and Device Technology IISB, Schottkystrasse 10, 91058 Erlangen, Germany
C
Christopher Straub
Department Modeling and Artificial Intelligence, Fraunhofer Institute for Integrated Systems and Device Technology IISB, Schottkystrasse 10, 91058 Erlangen, Germany
G
Georg Kruse
Department Modeling and Artificial Intelligence, Fraunhofer Institute for Integrated Systems and Device Technology IISB, Schottkystrasse 10, 91058 Erlangen, Germany
A
Andreas Rosskopf
Department Modeling and Artificial Intelligence, Fraunhofer Institute for Integrated Systems and Device Technology IISB, Schottkystrasse 10, 91058 Erlangen, Germany