arXiv subjects

Vlad Medvedev

Publications and source records attributed to Vlad Medvedev.

3 recordsLinked to original sources

arXiv2026

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

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.

cs.LG↗

arXiv2025

Physics-Informed PointNets for Modeling Electromagnetic Scattering from All-Dielectric Metasurfaces with Inclined Nanopillars

Metasurfaces are innovative planar optical structures capable of manipulating incident light properties. Accurate and computationally efficient modeling of such metasurfaces, particularly those with irregular geometries, remains a challenge for conventional solvers. In this work, we present a mesh-free Physics-Informed PointNet (PIPN) to model electromagnetic scattering from all-dielectric metasurfaces that feature spatially varying nanopillars. Our approach uses the PointNet architecture to directly encode spatially varying material properties into the Physics-Informed Machine Learning (PIML) framework. We demonstrate the generalization capability of our PIPN through evaluations on datasets; these datasets are generated with varying refractive indices representing common dielectric materials. Furthermore, the inclination angles are varied within each dataset, which represent expected manufacturing defects. Overall, our method provides a promising, mesh-free framework for accurate and efficient modeling of complex optical structures represented by irregular geometries.

physics.optics↗

arXiv2025

Hard-constraining Neumann boundary conditions in physics-informed neural networks via Fourier feature embeddings

We present a novel approach to hard-constrain Neumann boundary conditions in physics-informed neural networks (PINNs) using Fourier feature embeddings. Neumann boundary conditions are used to described critical processes in various application, yet they are more challenging to hard-constrain in PINNs than Dirichlet conditions. Our method employs specific Fourier feature embeddings to directly incorporate Neumann boundary conditions into the neural network's architecture instead of learning them. The embedding can be naturally extended by high frequency modes to better capture high frequency phenomena. We demonstrate the efficacy of our approach through experiments on a diffusion problem, for which our method outperforms existing hard-constraining methods and classical PINNs, particularly in multiscale and high frequency scenarios.

cs.LG↗