Search

Mitigating nonlinear weight update characteristics is one of the main challenges in designing neural networks based on synaptic devices. This paper presents a novel weight update method named conditional reverse update scheme (CRUS) for hardware neural network (HNN) consisting of synaptic devices with highly nonlinear or abrupt conductance update characteristics. We formulate a linear optimization method of conductance in synaptic devices to reduce the average deviation of weight changes from those calculated by the Stochastic Gradient Rule (SGD) algorithm. We introduce a metric called update noise (UN) to analyze the training dynamics during training. We then design a weight update rule that reduces the UN averaged over the training process. The optimized network achieves >90% accuracy on the MNIST dataset under highly nonlinear long-term potentiation (LTP) and long-term depression (LTD) conditions while using inaccurate and infrequent conductance sensing. Furthermore, the proposed method shows better accuracy than previously reported nonlinear weight update mitigation techniques under the same hardware specifications and device conditions. It also exhibits robustness to temporal variations in conductance updates. We expect our scheme to relieve design requirements in device and circuit engineering and serve as a practical technique that can be applied to future HNNs., Comment: 10 pages, 13 figures, 1 table

While the Helpman-Trajtenberg model is successful in explaining the macroeconomic cycle generated by GPT in the context of a closed economy, open-economy implications regarding GPT have received scant attention. We seek to fill such void in the previous literature by developing a dynamic North-South trade model with GPT. Our major findings are : 1) the introduction of GPT causes a product cycle from the North to the South followed by a reverse product cycle from the South to the North; 2) the presence of trading partner can cushion the macroeconomic shock experienced by the North.