Surrogate gradients for analog neuromorphic computing
Surrogate gradients for analog neuromorphic computing is a scholarly work, published in 2022 in ''Proceedings of the National Academy of Sciences of the United States of America''. The main subjects of the publication include computer architecture, reservoir computing, artificial intelligence, efficient energy use, computer science, artificial neural network, ferroelectric random-access memory, spike, memristor, neuromorphic engineering, energy, inference, and spiking neural network. The authors present a learning framework resulting in bioinspired spiking neural networks with high performance, low inference latency, and sparse spike-coding schemes, which also self-corrects for device mismatch.