arXiv subjects

Leon Armbruster

Publications and source records attributed to Leon Armbruster.

4 recordsLinked to original sources

arXiv2026

jNO: A JAX Library for Neural Operator and Foundation Model Training

jNO (jax Neural Operators) is a JAX-native library for neural operators and foundation models with unified support for both data-driven and physics-informed training. Its core design is a tracing system in which domains, model calls, residuals, supervised losses, and diagnostics are written in one symbolic language and compiled into one optimization pipeline. This allows users to move between operator regression, mesh-aware residual evaluation, and PDE-constrained training without restructuring the surrounding code. jNO also supports multi-model compositions, fine-grained control at parameter level (model, optimizer, and learning rate), hyperparameter tuning, and JAX-native workflows for translated PDE foundation-model families. The source repository is available at https://github.com/FhG-IISB/jNO.

cs.LG↗

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↗

arXiv2022

Nature of barriers determine first passage times in heterogeneous media

Intuition suggests that passage times across a region increases with the number of barriers along the path. Can this fail depending on the nature of the barrier? To probe this fundamental question, we exactly solve for the first passage time in general d-dimensions for diffusive transport through a spatially patterned array of obstacles - either entropic or energetic, depending on the nature of the obstacles. For energetic barriers, we show that first passage times vary non-monotonically with the number of barriers, while for entropic barriers it increases monotonically. This non-monotonicity for energetic barriers further reflects in the behaviour of effective diffusivity as well. We then design a simple experiment where a robotic bug navigates a heterogeneous environment through a spatially patterned array of obstacles to validate our predictions. Finally, using numerical simulations, we show that this non-monotonic behaviour for energetic barriers is general and extends to even super-diffusive transport.

cond-mat.stat-mech↗