English(EN) Correlation flow governs learning at criticality

新理论将相关性传播与深度学习中的神经切线核联系起来

作者 PulseAugur 编辑部 · [1 个来源] · 2026-08-11 04:00

研究人员在深度神经网络中的相关性传播与神经切线核（NTK）之间建立了理论联系。通过结合均值场理论和随机矩阵理论，他们证明了相关性传播到无限深度仅在权重-偏差方差平面上的临界点才可能发生。在该临界点，端到端雅可比行列式随着深度的增加而代数式地消失，导致NTK与输出相关性成正比。研究还表明，与高斯初始化相比，正交初始化能更好地控制有限宽度、有限深度的网络的渐近动力学。 AI

影响为理解深度学习初始化及其对信息传播和学习动力学的影响提供了理论框架。

排序理由详细介绍深度学习动力学理论进展的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.LG 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 0 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 详细介绍深度学习动力学理论进展的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: paper, other; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: 60 days old; Aged out of breaking-news scoring windows; ranking reflects the durable signal from the full source set.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv cs.LG TIER_1 English(EN) · Andrea Combette, Nelly Pustelnik, Antoine Venaille · 2026-08-11 04:00

相关流在临界点控制学习

arXiv:2608.08350v1 Announce Type: new Abstract: The initialisation of deep neural networks determines whether information and gradients can propagate across depth, yet a unified theory connecting these properties to learning dynamics remains elusive. Combining mean-field theory a…

本文如何被排名

报道来源 [1]

相关流在临界点控制学习

相关实体

相关话题