Saved in:
Bibliographic Details
Main Authors: Cao, Sansheng, Ma, Zhengyu, Tian, Yonghong
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2602.10607
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910018839773184
author Cao, Sansheng
Ma, Zhengyu
Tian, Yonghong
author_facet Cao, Sansheng
Ma, Zhengyu
Tian, Yonghong
contents Zeroth-order (ZO) optimization has long been favored for its biological plausibility and its capacity to handle non-differentiable objectives, yet its computational complexity has historically limited its application in deep neural networks. Challenging the conventional paradigm that gradients propagate layer-by-layer, we propose Hierarchical Zeroth-Order (HZO) optimization, a novel divide-and-conquer strategy that decomposes the depth dimension of the network. We prove that HZO reduces the query complexity from $O(ML^2)$ to $O(ML \log L)$ for a network of width $M$ and depth $L$, representing a significant leap over existing ZO methodologies. Furthermore, we provide a detailed error analysis showing that HZO maintains numerical stability by operating near the unitary limit ($L_{lip} \approx 1$). Extensive evaluations on CIFAR-10 and ImageNet demonstrate that HZO achieves competitive accuracy compared to backpropagation.
format Preprint
id arxiv_https___arxiv_org_abs_2602_10607
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Hierarchical Zero-Order Optimization for Deep Neural Networks
Cao, Sansheng
Ma, Zhengyu
Tian, Yonghong
Machine Learning
Artificial Intelligence
Zeroth-order (ZO) optimization has long been favored for its biological plausibility and its capacity to handle non-differentiable objectives, yet its computational complexity has historically limited its application in deep neural networks. Challenging the conventional paradigm that gradients propagate layer-by-layer, we propose Hierarchical Zeroth-Order (HZO) optimization, a novel divide-and-conquer strategy that decomposes the depth dimension of the network. We prove that HZO reduces the query complexity from $O(ML^2)$ to $O(ML \log L)$ for a network of width $M$ and depth $L$, representing a significant leap over existing ZO methodologies. Furthermore, we provide a detailed error analysis showing that HZO maintains numerical stability by operating near the unitary limit ($L_{lip} \approx 1$). Extensive evaluations on CIFAR-10 and ImageNet demonstrate that HZO achieves competitive accuracy compared to backpropagation.
title Hierarchical Zero-Order Optimization for Deep Neural Networks
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2602.10607