Low-power analogue neural networks with trainable nonlinear connections for continuous control
具有可训练非线性连接的低功耗模拟神经网络用于连续控制
机构 * School of Computer Science, University of Sheffield(谢菲尔德大学计算机科学学院) ; Intrinsic Semiconductor Technologies(Intrinsic Semiconductor Technologies公司) ; School of Chemical, Biological, and Materials Science Engineering, University of Sheffield(谢菲尔德大学化学、生物与材料科学工程学院) ; School of Physics, Engineering, and Technology, University of York(约克大学物理、工程与技术学院) ; Department of Computer Science, University of York(约克大学计算机科学系) ; Department of Electronic & Electrical Engineering, University College London(伦敦大学学院电子与电气工程系) ; King’s College London(伦敦国王学院) ; Blackett Laboratory, Imperial College London(帝国理工学院布莱克特实验室)
专题命中 机器人数据与评测 :robotic(abstract);分类 cs.AI、cs.LG
AI总结 受Kolmogorov-Arnold网络启发,在连接上放置可训练非线性函数,使每个物理连接成为可学习计算单元,通过现场可编程模拟阵列实现带通滤波器,在连续控制等任务上以更少节点和连接达到高效,预计CMOS实现功耗约30微瓦。
Comments Preprint. Further verification of all simulations is ongoing. Any resulting corrections will be incorporated in a revised version