Effect of layer-specific synaptic retention characteristics on the accuracy of deep neural networks
Solid-State Electronics(2023)
摘要
•The effect of retention loss on neuromorphic synaptic devices is investigated.•Depending on the types of activation functions or types of neural network layers, the effects of weight/bias loss of synaptic devices are different.•By adjusting the bias of the FC layer, the performance and retention characteristics of the neural network can be improved.
更多查看译文
关键词
Charge trap device,Data retention reliability,Neuromorphic computing
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要