发表机构
Federal University of Bahia; Grupo de Estudos e Aplicação de Inteligência Artificial em Geofísica (GAIA), Federal University of Bahia（巴伊亚联邦大学; 巴伊亚联邦大学地球物理学人工智能研究与应用组（GAIA））
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本文提出SILVA网络，将其作为结构化隐式层与向量吸引子，通过动态交互场分离多类影响，经多领域基准实验验证其动态可训练、可诊断的隐式表示能力。
AI 中文摘要
许多学习问题需要能协调直接输入、邻近结构与更广上下文的表示。在隐式神经层中，这些影响通常被吸收为单一的不动点更新，难以区分哪些来自刺激、哪些局部传播、哪些来自全局上下文、哪些由求解器动态产生。本文引入SILVA网络（Structured Implicit Layers and Vector Attractors via Dynamic Interaction Fields），其在一个不动点架构内分离了刺激、局部交互、全局交互、阻尼与读出操作。该架构可通过节点、邻域与全局摘要的领域特定定义，实例化用于图像、分子、引文网络及长程图基准任务。实验与 ablation 显示这些项具有任务依赖作用：局部交互在图任务中起关键作用，MNIST在测试规模下几乎未从循环中获益，长程节点分类基准中全局收益最显著。因此，SILVA提供的隐式表示其内部交互动态可被训练、 ablation、可视化与诊断。
英文摘要
Many learning problems require representations that reconcile direct input, nearby structure, and broader context. In implicit neural layers, these influences are usually absorbed into a single fixed-point update, making it hard to identify what enters from the stimulus, what propagates locally, what comes from global context, and what is produced by solver dynamics. Here we introduce SILVA Networks, Structured Implicit Layers and Vector Attractors via Dynamic Interaction Fields. SILVA separates stimulus, local interaction, global interaction, damping, and readout inside one fixed-point architecture. The same template is instantiated for images, molecules, citation networks, and long-range graph benchmarks through domain-specific definitions of nodes, neighborhoods, and global summaries. Experiments and ablations show task-dependent roles for these terms: local interactions are load-bearing in the graph tasks, MNIST gains little from recurrence at the tested capacity, and the clearest global benefit appears in a long-range node-classification benchmark. SILVA therefore provides an implicit representation whose internal interaction dynamics can be trained, ablated, visualized, and diagnosed.
Comments46 pages, 10 figures
