发表机构
Department of Computer Science and Communication（计算机科学与通信系）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
提出MetaLearnNCA，利用交互式神经细胞自动机实现无梯度的少样本元学习，在保持分布内性能的同时提升分布外迁移效果。
AI 中文摘要
少样本元学习传统上将任务适应形式化为通过展开计算图的解析梯度下降，或基于扁平化一维特征向量的度量距离比较，前者在测试时产生昂贵的反向传播，后者则丢弃了固有的二维空间几何信息。在这项工作中，我们提出了METALEARNNCA，一个去中心化框架，通过耦合神经细胞自动机（NCA）的动态交互实现少样本适应，在推理过程中无需计算解析梯度。MetaLearnNCA将任务适应分解为Active-NCA，它基于称为空间程序的连续二维空间记忆网格执行任务推理，以及一个学习的Meta-NCA，它作为去中心化的细胞优化器，通过跨局部邻域扩散空间误差残差来动态更新该程序。METALEARN-NCA在分布内与经典元学习器相比具有竞争力（Omniglot上96.12%），并在MNIST、KMNIST和Fashion-MNIST迁移中，跨10个独立测试种子、1-shot、5-shot和10-shot设置下取得了分布外迁移增益（例如，在10-shot MNIST上超过原型网络+10.54%，在10-shot Fashion-MNIST上超过FOMAML +3.87%）。我们的结果表明，稳健的、无梯度的学习如何学习可以从非冯·诺依曼基底上的去中心化细胞动力学中涌现。
英文摘要
Few-shot meta-learning traditionally formulates task adaptation either as analytical gradient descent through unrolled computational graphs or as metric-based distance comparisons over flattened 1D fea- ture vectors, which either incur costly test-time backpropagation or discard native 2D spatial geometry. In this work, we propose METALEARNNCA, a decentralized framework that achieves few-shot adapta- tion through the dynamical interaction of coupled Neural Cellular Automata (NCAs) without computing analytical gradients during inference. MetaLearnNCA decomposes task adaptation into an Active- NCA, which executes task inference conditioned on a continuous 2D spatial memory grid termed the spatial program, and a learned Meta-NCA, which acts as a decentralized cellular optimizer by diffusing spatial error residuals across local neighborhoods to dynamically update this program. METALEARN- NCA is competitive against canonical meta-learners in-distribution (96.12% on Omniglot) with Out-Of- Distribution transfer gains on MNIST, KMNIST, and Fashion-MNIST transfer across 10 independent testing seeds across 1-, 5-, and 10-shot regimes (e.g., surpassing Prototypical Networks by +10.54% on 10-shot MNIST and a +3.87% gain on 10-shot Fashion-MNIST over FOMAML). Our results establish that robust, gradient-free learning-to-learn can emerge from decentralized cellular dynamics on non-von Neumann substrates.
