English

Born-Infeld (BI) for AI: Energy-Conserving Descent (ECD) for Optimization

Machine Learning 2023-05-16 v2 Cosmology and Nongalactic Astrophysics High Energy Physics - Theory Optimization and Control Machine Learning

Abstract

We introduce a novel framework for optimization based on energy-conserving Hamiltonian dynamics in a strongly mixing (chaotic) regime and establish its key properties analytically and numerically. The prototype is a discretization of Born-Infeld dynamics, with a squared relativistic speed limit depending on the objective function. This class of frictionless, energy-conserving optimizers proceeds unobstructed until slowing naturally near the minimal loss, which dominates the phase space volume of the system. Building from studies of chaotic systems such as dynamical billiards, we formulate a specific algorithm with good performance on machine learning and PDE-solving tasks, including generalization. It cannot stop at a high local minimum, an advantage in non-convex loss functions, and proceeds faster than GD+momentum in shallow valleys.

Keywords

Cite

@article{arxiv.2201.11137,
  title  = {Born-Infeld (BI) for AI: Energy-Conserving Descent (ECD) for Optimization},
  author = {G. Bruno De Luca and Eva Silverstein},
  journal= {arXiv preprint arXiv:2201.11137},
  year   = {2023}
}

Comments

ICML 2022. 9 pages + Appendix, 8 figures. Code available at https://github.com/gbdl/BBI

R2 v1 2026-06-24T09:04:18.762Z