用脑启发的注意力机制提升神经元自动机的自组织能力
BraiNCA: brain-inspired neural cellular automata and applications to morphogenesis and motor control
- 引入注意力层与长程连接，模拟大脑非局部连通性
- 在形态发生与运动控制任务中学习速度更快、抗损伤更强
- 适合研究生物真实网络结构下的集体计算机制
现有神经细胞自动机（NCAs）多基于规则网格与局部邻域（Moore邻域），缺乏长程连接与复杂拓扑。本文提出脑启发的BraiNCA，引入注意力机制、长程连接与复杂拓扑结构。实验表明，相比传统NCAs，BraiNCA在形态发生与运动控制任务中表现出更强的鲁棒性与更快的学习速度，证明结合注意力消息选择与显式长程边可实现更高效的样本利用与损伤容错性。结果支持假设：对于需要跨空间与时间尺度协调的任务，交互拓扑结构与动态信息路由能力显著影响学习效率。BraiNCA在保持去中心化局部更新原则的同时，更贴近真实大脑的非局部连接模式，为研究生物合理网络结构下的集体计算与演化认知基质提供了有前景的框架。
原文摘要 · Abstract (English)
Most of the Neural Cellular Automata (NCAs) defined in the literature have a common theme: they are based on regular grids with a Moore neighborhood (one-hop neighbour). They do not take into account long-range connections and more complex topologies as we can find in the brain. In this paper, we introduce BraiNCA, a brain-inspired NCA with an attention layer, long-range connections and complex topology. BraiNCAs shows better results in terms of robustness and speed of learning on the two tasks compared to Vanilla NCAs establishing that incorporating attention-based message selection together with explicit long-range edges can yield more sample-efficient and damage-tolerant self-organization than purely local, grid-based update rules. These results support the hypothesis that, for tasks requiring distributed coordination over extended spatial and temporal scales, the choice of interaction topology and the ability to dynamically route information will impact the robustness and speed of learning of an NCA. More broadly, BraiNCA provides brain-inspired NCA formulation that preserves the decentralized local update principle while better reflecting non-local connectivity patterns, making it a promising substrate for studying collective computation under biologically-realistic network structure and evolving cognitive substrates.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。