一种能够在片上学习的原子玻尔兹曼机
An atomic Boltzmann machine capable of on-chip learning
摘要：
玻尔兹曼机（Boltzmann Machine，BM）是一种由随机放电神经元组成的神经网络，它可以通过调整神经元之间的突触相互作用来学习复杂的概率分布。BM代表了一类非常通用的随机神经网络，可用于数据聚类、生成建模和深度学习。基于软件的随机神经网络的一个主要缺点是所需的Monte Carlo采样，其随着神经元的数量而难以扩展。在这里，我们实现了一个物理实现的BM直接在随机自旋动力学的半导体黑磷表面上的耦合钴原子的门控系综。实施轨道存储器的概念，利用扫描隧道显微镜，我们证明了自下而上的原子系综的随机电流噪声是由一个可重构的多阱能量景观定义的建设。利用黑磷的各向异性行为，我们构建了具有两个分离良好的内在时间尺度的原子系综，代表神经元和突触。通过表征给定突触配置的神经元的条件稳态分布，我们说明了一个合奏可以代表许多不同的概率分布。通过探索内在的突触动力学，我们揭示了一个自主重组的突触响应外部电刺激。这种自适应架构为直接在原子级机器学习硬件中进行片上学习铺平了道路。
英文摘要：
The Boltzmann Machine (BM) is a neural network composed of stochastically firing neurons that can learn complex probability distributions by adapting the synaptic interactions between the neurons. BMs represent a very generic class of stochastic neural networks that can be used for data clustering, generative modelling and deep learning. A key drawback of software-based stochastic neural networks is the required Monte Carlo sampling, which scales intractably with the number of neurons. Here, we realize a physical implementation of a BM directly in the stochastic spin dynamics of a gated ensemble of coupled cobalt atoms on the surface of semiconducting black phosphorus. Implementing the concept of orbital memory utilizing scanning tunnelling microscopy, we demonstrate the bottom-up construction of atomic ensembles whose stochastic current noise is defined by a reconfigurable multi-well energy landscape. Exploiting the anisotropic behaviour of black phosphorus, we build ensembles of atoms with two well-separated intrinsic time scales that represent neurons and synapses. By characterizing the conditional steady-state distribution of the neurons for given synaptic configurations, we illustrate that an ensemble can represent many distinct probability distributions. By probing the intrinsic synaptic dynamics, we reveal an autonomous reorganization of the synapses in response to external electrical stimuli. This self-adaptive architecture paves the way for on-chip learning directly in atomic-scale machine learning hardware.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
