PulseAugur
中
实时 21:42:37
English(EN) Recurrent Graph Neural Networks and Arithmetic Circuits

研究人员探索神经网络的复杂性、计算和图论的联系

作者 PulseAugur 编辑部 · [4 个来源] ·

研究人员正在探索神经网络的新理论框架和计算模型。一篇论文通过对张量运算建模，引入了一个统一的框架来分析和构建深度神经网络，揭示了历史上的架构复杂性趋势，并识别了未被探索的高复杂度架构。另一项研究将动力系统和图论统一起来，以理解循环神经网络中的计算，并提出了 resolvent-RNNs，它约束多跳路径以提高时间稀疏性和性能。第三篇论文建立了循环图神经网络的表达能力与循环算术电路之间的精确对应关系，从电路复杂性理论提供了新的视角。 AI

影响 这些理论上的进步可能带来更高效、更强大的神经网络架构，并加深对其计算机制的理解。

排序理由 多篇 arXiv 论文发表了关于神经网络的理论框架和计算模型。

在 arXiv cs.LG 阅读 →

AI 生成摘要 · Google Gemini · 来自 4 个来源。 我们如何撰写摘要 →

研究人员探索神经网络的复杂性、计算和图论的联系

报道来源 [4]

  1. arXiv cs.LG TIER_1 English(EN) · Nicholas J. Cooper, Fran\c{c}ois G. Meyer, Michael L. Roberts, Carlos Zapata-Carratal\'a, Lijun Chen, Danna Gurari ·

    On the Architectural Complexity of Neural Networks

    arXiv:2605.04325v1 Announce Type: new Abstract: We introduce a unified theoretical framework for the rigorous analysis and systematic construction of deep neural networks (DNNs). This framework addresses a gap in existing theory by explicitly modeling the structure of tensor oper…

  2. arXiv cs.AI TIER_1 English(EN) · Jatin Sharma, Dan F. M Goodman, Danyal Akarca ·

    Unifying Dynamical Systems and Graph Theory to Mechanistically Understand Computation in Neural Networks

    arXiv:2605.03598v2 Announce Type: cross Abstract: Understanding how biological and artificial neural networks implement computation from connectivity is a central problem in neuroscience and machine learning. In neural systems, structural and functional connectivity are known to …

  3. arXiv cs.AI TIER_1 English(EN) · Dan F. M Goodman ·

    Unifying Dynamical Systems and Graph Theory to Mechanistically Understand Computation in Neural Networks

    Understanding how biological and artificial neural networks implement computation from connectivity is a central problem in neuroscience and machine learning. In neural systems, structural and functional connectivity are known to diverge, motivating approaches that move beyond di…

  4. arXiv cs.LG TIER_1 English(EN) · Timon Barlag, Vivian Holzapfel, Laura Strieker, Jonni Virtema, Heribert Vollmer ·

    Recurrent Graph Neural Networks and Arithmetic Circuits

    arXiv:2603.05140v2 Announce Type: replace-cross Abstract: We characterise the computational power of recurrent graph neural networks (GNNs) in terms of arithmetic circuits over the real numbers. Our networks are not restricted to aggregate-combine GNNs or other particular types. …

相关实体

相关话题