AI 中文总结
研究用光学自旋量子比特升级量子-经典神经元，通过实验测量集成后的神经元，推导考虑其与量子比特自旋相互作用的量子模型并计算，以预测量子计算操作在该系统状态空间的扩展，对实现有量子光学能力的硬件神经元至关重要。
AI 中文摘要
包含内置存储组件的硬件神经元是实现大规模神经网络的关键技术，这些网络表现出自适应巡回行为并产生与神经科学已知操作相匹配的生物启发式尖峰模式。能够通过其量子比特状态和量子轨迹进行全量子信息处理的神经元，通过相干/纠缠和非马尔可夫性水平提供了混合量子-经典操作的能力。现在用近红外脉冲激光激发的光学自旋量子比特升级这些硬件神经元，由半导体碳化硅中的带负电硅空位和超导铌/氧化铌薄膜中的钕稀土离子设计形成存储器，可获得考虑电子和核相互作用的独特自旋哈密顿量。在本文中，我们通过实验测量了与光学自旋量子比特集成的光电量子-经典神经元，并研究了定制的神经元尖峰序列如何驱动光致发光振荡并调制量子自旋跃迁。我们推导了一个考虑神经元与光学量子比特自旋之间相互作用的系统量子模型，并进行计算以预测量子计算操作如何在量子-经典神经元和自旋量子比特系统的状态空间中扩展。例如，由于动态内存变化，强度可调对神经元量子态产生的影响。这些结果对于实现具有量子光学能力且受生物体内生物光子存在启发的硬件神经元至关重要。
英文摘要
Hardware neurons incorporating built-in memory components are critical technologies for implementing large scale neural networks that exhibit adaptive-itinerant behavior and produce biological-inspired spiking patterns that match neuroscience known operations (1). Moreover, neurons capable of full quantum information processing through their qubit states and quantum trajectory provides capability for hybrid quantum-classical operations through the level of coherence/entanglement and non-Markovianity (2,3). These attributes further contribute to solving complex quantum mechanical problems by providing a rich and diverse group for expressing quantum neural states. Upgrading these hardware neurons now with optical spin qubits excitable with pulsed laser excitation in the near infrared and designed from negatively charged silicon vacancies in semiconductor silicon carbide (4) and Neodymium rare earth ions in the superconducting Niobium/Niobium oxide film (5) that form the memory provides access to the unique spin Hamiltonian that considers electronic and nuclear interactions. In this manuscript, we experimentally measure opto-electronic quantum-classical neurons integrated with a optical spin qubit and examine how tailored neuronal spiking sequences drive the photoluminescence oscillations and modulate the quantum spin transitions. We derive a quantum model for the system that considers the interaction between the neuron and optical qubit spin and perform calculations to project how quantum computing operations are expandable across the state space in the quantum-classical neuron and spin qubit system. For example, the resulting impact on the neuron quantum states where the strength is adjustable due to dynamical memory changes. These results are critical for realizing hardware neurons with quantum optical capabilities and inspired by the existence in biology of biophotons.
