Neural ODEs with Invariants by Construction

Read full text Summarise with VIRAT
Authors: Fangzhou Yu, Yiqi Su, Ray Lee, Shenfeng Cheng, Naren Ramakrishnan
Cited by: 0
References: 51

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Neural ODEs are increasingly used as continuous-time models for scientific and sensor data, but unconstrained neural ODEs can drift and violate domain invariants (e.g., conservation laws), yielding physically implausible solutions. In turn, this can compound error in long-horizon prediction and surrogate simulation. Existing solutions typically aim to enforce invari- ance by soft penalties or other forms of regularization, which can reduce overall error but do not guarantee that trajectories will not leave the constraint manifold. We introduce a frame- work that enforces invariants by construction: it treats invari- ants as first-class types and uses an LLM-driven compilation workflow to translate a generic neural ODE specification into a structure-preserving architecture whose trajectories remain on the admissible manifold in continuous time(and up to numer- ical integration error in practice). This compiler view cleanly separates what must be preserved (scientific structure) from what is learned from data (dynamics within that structure). It provides a systematic design pattern for invariant-respecting neural surrogates across scientific domains.

Fields of Study & Subjects

Model Reduction and Neural Networks Numerical Solvers for Ordinary Differential Equations Neural Networks and Applications

Publication details

Journal / source
Proceedings of the AAAI Symposium Series
Volume
10
Issue
1
Pages
695–702
Publication date
6 November 2026
Language
EN