论文标题
神经形态系统的可塑性
Percolation with plasticity for neuromorphic systems
论文作者
论文摘要
我们通过可塑性系统（PWPS）渲染神经形态计算感兴趣的特性发展了渗透理论。与两个大电极之间的标准渗透不同，它们具有多个（$ n \ gg 1 $）接口，并且它们之间的导电途径呈指数级数（$ n！$）。这些途径由非OHMIC随机电阻组成，这些电阻会导致偏置引起的非易失性修饰（可塑性）。 PWPS的神经形态特性包括：能够将输入数据转换为时空模式的多价值内存，高维度和非线性，可调性褪色的内存确保更多取决于最近输入的输出，并且不需要大量互连。这里功能的一些概念示例是随机数生成，矩阵向量乘法和关联内存。了解PWP拓扑，统计和操作为自己的进一步的理论和实验见解打开了一个领域。
We develop a theory of percolation with plasticity systems (PWPs) rendering properties of interest for neuromorphic computing. Unlike the standard percolation between two large electrodes, they have multiple ($N\gg 1$) interfaces and exponentially large number ($N!$) of conductive pathways between them. These pathways consist of non-ohmic random resistors that can undergo bias induced nonvolatile modifications (plasticity). The neuromorphic properties of PWPs include: multi-valued memory, high dimensionality and nonlinearity capable of transforming input data into spatiotemporal patterns, tunably fading memory ensuring outputs that depend more on recent inputs, and no need for massive interconnects. A few conceptual examples of functionality here are random number generation, matrix-vector multiplication, and associative memory. Understanding PWP topology, statistics, and operations opens a field of its own calling upon further theoretical and experimental insights.
