English

Correcting Auto-Differentiation in Neural-ODE Training

Machine Learning 2026-03-31 v3 Numerical Analysis Numerical Analysis

Abstract

Does the use of auto-differentiation yield reasonable updates for deep neural networks (DNNs)? Specifically, when DNNs are designed to adhere to neural ODE architectures, can we trust the gradients provided by auto-differentiation? Through mathematical analysis and numerical evidence, we demonstrate that when neural networks employ high-order methods, such as Linear Multistep Methods (LMM) or Explicit Runge-Kutta Methods (ERK), to approximate the underlying ODE flows, brute-force auto-differentiation often introduces artificial oscillations in the gradients that prevent convergence. In the case of Leapfrog and 2-stage ERK, we propose simple post-processing techniques that effectively eliminates these oscillations, correct the gradient computation and thus returns the accurate updates.

Keywords

Cite

@article{arxiv.2306.02192,
  title  = {Correcting Auto-Differentiation in Neural-ODE Training},
  author = {Yewei Xu and Shi Chen and Qin Li},
  journal= {arXiv preprint arXiv:2306.02192},
  year   = {2026}
}

Comments

Accepted for publication in SIAM Journal on Applied Mathematics. This version corresponds to the final draft, prior to copyediting and production

R2 v1 2026-06-28T10:55:34.669Z