TINNs: Time-Induced Neural Networks for Solving Time-Dependent PDEs

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Dai, Chen-Yang, Chang, Che-Chia, Lin, Te-Sheng, Lai, Ming-Chih, Lai, Chieh-Hsin
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917227168530432
author Dai, Chen-Yang
Chang, Che-Chia
Lin, Te-Sheng
Lai, Ming-Chih
Lai, Chieh-Hsin
author_facet Dai, Chen-Yang
Chang, Che-Chia
Lin, Te-Sheng
Lai, Ming-Chih
Lai, Chieh-Hsin
contents Physics-informed neural networks (PINNs) solve time-dependent partial differential equations (PDEs) by learning a mesh-free, differentiable solution that can be evaluated anywhere in space and time. However, standard space--time PINNs take time as an input but reuse a single network with shared weights across all times, forcing the same features to represent markedly different dynamics. This coupling degrades accuracy and can destabilize training when enforcing PDE, boundary, and initial constraints jointly. We propose Time-Induced Neural Networks (TINNs), a novel architecture that parameterizes the network weights as a learned function of time, allowing the effective spatial representation to evolve over time while maintaining shared structure. The resulting formulation naturally yields a nonlinear least-squares problem, which we optimize efficiently using a Levenberg--Marquardt method. Experiments on various time-dependent PDEs show up to $4\times$ improved accuracy and $10\times$ faster convergence compared to PINNs and strong baselines.
format Preprint
id arxiv_https___arxiv_org_abs_2601_20361
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle TINNs: Time-Induced Neural Networks for Solving Time-Dependent PDEs
Dai, Chen-Yang
Chang, Che-Chia
Lin, Te-Sheng
Lai, Ming-Chih
Lai, Chieh-Hsin
Machine Learning
Numerical Analysis
Physics-informed neural networks (PINNs) solve time-dependent partial differential equations (PDEs) by learning a mesh-free, differentiable solution that can be evaluated anywhere in space and time. However, standard space--time PINNs take time as an input but reuse a single network with shared weights across all times, forcing the same features to represent markedly different dynamics. This coupling degrades accuracy and can destabilize training when enforcing PDE, boundary, and initial constraints jointly. We propose Time-Induced Neural Networks (TINNs), a novel architecture that parameterizes the network weights as a learned function of time, allowing the effective spatial representation to evolve over time while maintaining shared structure. The resulting formulation naturally yields a nonlinear least-squares problem, which we optimize efficiently using a Levenberg--Marquardt method. Experiments on various time-dependent PDEs show up to $4\times$ improved accuracy and $10\times$ faster convergence compared to PINNs and strong baselines.
title TINNs: Time-Induced Neural Networks for Solving Time-Dependent PDEs
topic Machine Learning
Numerical Analysis
url https://arxiv.org/abs/2601.20361