English(EN) Hidden Activations are not Enough I: Knowledge Matrices as Higher Representations

新的“知识矩阵”为神经网络提供更高阶表示

作者 PulseAugur 编辑部 · [1 个来源] · 2026-09-28 02:43

本文介绍了“知识矩阵”，这是一种在训练过的前馈神经网络中表示信息的创新方法。这些矩阵源自网络权重和激活，提供了比传统隐藏激活更高阶的表示。研究表明，知识矩阵可用于分析网络行为，测量ResNet-152和DenseNet-121等不同架构之间的距离，甚至量化对抗性攻击的影响。 AI

影响引入了理解和比较神经网络表示的新理论框架。

排序理由介绍用于分析神经网络新理论概念的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]

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; 介绍用于分析神经网络新理论概念的学术论文。[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, model release; 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: 5 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。

arXiv cs.NE (Neural & Evolutionary) TIER_1 English(EN) · Marco Armenta · 2026-09-28 02:43

隐藏激活不足以 I：知识矩阵作为更高阶表示

We study the knowledge matrix of a trained feedforward network as a higher representation of its inputs. A network is a pair $(W,f)$, a thin representation $W$ of its quiver and an activation $f$; its function factorizes through the space of quiver representations, each input $x$…