English

Towards Interpretable Deep Local Learning with Successive Gradient Reconciliation

Machine Learning 2024-06-11 v1 Neural and Evolutionary Computing

Abstract

Relieving the reliance of neural network training on a global back-propagation (BP) has emerged as a notable research topic due to the biological implausibility and huge memory consumption caused by BP. Among the existing solutions, local learning optimizes gradient-isolated modules of a neural network with local errors and has been proved to be effective even on large-scale datasets. However, the reconciliation among local errors has never been investigated. In this paper, we first theoretically study non-greedy layer-wise training and show that the convergence cannot be assured when the local gradient in a module w.r.t. its input is not reconciled with the local gradient in the previous module w.r.t. its output. Inspired by the theoretical result, we further propose a local training strategy that successively regularizes the gradient reconciliation between neighboring modules without breaking gradient isolation or introducing any learnable parameters. Our method can be integrated into both local-BP and BP-free settings. In experiments, we achieve significant performance improvements compared to previous methods. Particularly, our method for CNN and Transformer architectures on ImageNet is able to attain a competitive performance with global BP, saving more than 40% memory consumption.

Keywords

Cite

@article{arxiv.2406.05222,
  title  = {Towards Interpretable Deep Local Learning with Successive Gradient Reconciliation},
  author = {Yibo Yang and Xiaojie Li and Motasem Alfarra and Hasan Hammoud and Adel Bibi and Philip Torr and Bernard Ghanem},
  journal= {arXiv preprint arXiv:2406.05222},
  year   = {2024}
}

Comments

ICML 2024

R2 v1 2026-06-28T16:57:48.138Z