In-context modeling as a retrain-free paradigm for foundation models in computational science

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Li, Lingfeng, Li, Zhuoyuan, Li, Shun, Zhan, Kaixin, Gao, Huajian, Chen, Changqing, Yang, Liu
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866918467618209792
author Li, Lingfeng
Li, Zhuoyuan
Li, Shun
Zhan, Kaixin
Gao, Huajian
Chen, Changqing
Yang, Liu
author_facet Li, Lingfeng
Li, Zhuoyuan
Li, Shun
Zhan, Kaixin
Gao, Huajian
Chen, Changqing
Yang, Liu
contents Building models that generalize across physical systems without retraining remains a central challenge in computational science. Here we introduce In-Context Modeling (ICM), a retrain-free paradigm that infers physical relationships directly from observational fields. Rather than encoding system-specific behavior in fixed parameters, ICM assimilates measurements as physical context and performs inference through a single forward pass. Trained in a physics-informed, label-free manner using governing equations, a single model generalizes across unseen materials, geometries, and loading conditions. Demonstrated on hyperelasticity, ICM integrates with finite-element simulations and is validated using experimental full-field measurements. Moreover, performance improves with increasing data diversity and computational budget, exhibiting favorable scaling behavior analogous to foundation models. By recasting physical modeling as in-context inference, this work establishes a transferable paradigm for retrain-free scientific learning and a foundation for scalable modeling across computational science.
format Preprint
id arxiv_https___arxiv_org_abs_2604_23098
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle In-context modeling as a retrain-free paradigm for foundation models in computational science
Li, Lingfeng
Li, Zhuoyuan
Li, Shun
Zhan, Kaixin
Gao, Huajian
Chen, Changqing
Yang, Liu
Computational Engineering, Finance, and Science
Building models that generalize across physical systems without retraining remains a central challenge in computational science. Here we introduce In-Context Modeling (ICM), a retrain-free paradigm that infers physical relationships directly from observational fields. Rather than encoding system-specific behavior in fixed parameters, ICM assimilates measurements as physical context and performs inference through a single forward pass. Trained in a physics-informed, label-free manner using governing equations, a single model generalizes across unseen materials, geometries, and loading conditions. Demonstrated on hyperelasticity, ICM integrates with finite-element simulations and is validated using experimental full-field measurements. Moreover, performance improves with increasing data diversity and computational budget, exhibiting favorable scaling behavior analogous to foundation models. By recasting physical modeling as in-context inference, this work establishes a transferable paradigm for retrain-free scientific learning and a foundation for scalable modeling across computational science.
title In-context modeling as a retrain-free paradigm for foundation models in computational science
topic Computational Engineering, Finance, and Science
url https://arxiv.org/abs/2604.23098