Jha Aashu, Huang Chaoran, Prucnal Paul R
Opt Lett. 2020 Sep 1;45(17):4819-4822. doi: 10.1364/OL.398234.
We experimentally demonstrate all-optical reconfigurable nonlinear activation functions in a cavity-loaded Mach-Zehnder interferometer device on a silicon photonics platform, via the free-carrier dispersion effect. Our device is programmable to generate various nonlinear activation functions, including sigmoid, radial-basis, clamped rectified linear unit, and softplus, with tunable thresholds. We simulate benchmark tasks such as XOR and MNIST handwritten digit classifications with experimentally measured activation functions and obtain accuracies of 100% and 94%, respectively. Our device can serve as nonlinear units in photonic neural networks, while its nonlinear transfer function can be flexibly programmed to optimize the performance of different neuromorphic tasks.
我们通过自由载流子色散效应，在硅光子平台上的腔加载马赫-曾德尔干涉仪装置中，通过实验证明了全光可重构非线性激活函数。我们的装置可编程生成各种非线性激活函数，包括 sigmoid、径向基、钳位整流线性单元和 softplus，且阈值可调。我们使用实验测量的激活函数模拟了诸如异或和 MNIST 手写数字分类等基准任务，分别获得了 100%和 94%的准确率。我们的装置可作为光子神经网络中的非线性单元，同时其非线性传递函数可灵活编程以优化不同神经形态任务的性能。
