Our latest collaborative work with Deji Akinwande has just been published!
We combine graphene and Nafion to create a fully bio-compatible artificial synapse and show it has low update energy, good endurance, and good nonlinearity and symmetry for neural network applications. We show the metaplastic properties of the synapse can allow them to outperform ideal synapses in online learning neuromorphic tasks. Perhaps biological and artificial neural networks will be working in tandem, sooner rather than later!
Big shout-out to the co-first authors Dmitry Kireev and Sam Liu, and co-authors Harrison Jin, Patrick Xiao, and Christopher H. Bennett. And I want to acknowledge the research sponsors, this project brought together researchers supported by various programs, including the UT Austin NSF MRSEC, National Science Foundation (NSF), Sandia National Laboratories, Office of Naval Research, and the UT Austin Portugal Program.
Click here to read the article!
