English

Learning Discretized Neural Networks under Ricci Flow

Machine Learning 2024-12-20 v5 Information Theory Neural and Evolutionary Computing math.IT

Abstract

In this paper, we study Discretized Neural Networks (DNNs) composed of low-precision weights and activations, which suffer from either infinite or zero gradients due to the non-differentiable discrete function during training. Most training-based DNNs in such scenarios employ the standard Straight-Through Estimator (STE) to approximate the gradient w.r.t. discrete values. However, the use of STE introduces the problem of gradient mismatch, arising from perturbations in the approximated gradient. To address this problem, this paper reveals that this mismatch can be interpreted as a metric perturbation in a Riemannian manifold, viewed through the lens of duality theory. Building on information geometry, we construct the Linearly Nearly Euclidean (LNE) manifold for DNNs, providing a background for addressing perturbations. By introducing a partial differential equation on metrics, i.e., the Ricci flow, we establish the dynamical stability and convergence of the LNE metric with the L2L^2-norm perturbation. In contrast to previous perturbation theories with convergence rates in fractional powers, the metric perturbation under the Ricci flow exhibits exponential decay in the LNE manifold. Experimental results across various datasets demonstrate that our method achieves superior and more stable performance for DNNs compared to other representative training-based methods.

Keywords

Cite

@article{arxiv.2302.03390,
  title  = {Learning Discretized Neural Networks under Ricci Flow},
  author = {Jun Chen and Hanwen Chen and Mengmeng Wang and Guang Dai and Ivor W. Tsang and Yong Liu},
  journal= {arXiv preprint arXiv:2302.03390},
  year   = {2024}
}
R2 v1 2026-06-28T08:33:58.338Z