Transient memory and learning in correlated oxide neuromorphic devices

Biological neural systems can learn and forget information that is one possible mechanism for stability and lifelong learning of neural circuits. Emulating such features in electronic devices is essential for advancing neuromorphic electronics. We discuss examples of memory devices using strongly correlated oxides to illustrate learning behavior in this conference proceeding. We give examples of transient memory and forgetting dynamics by controlling the strength of the electrical stimuli as well as stochastic behavior. Using examples of prototypical Mott insulators such as NiO and VO2, we present our vision for a neuromorphic platform utilizing quantum materials. The studies inform design of electronic hardware in emerging AI and can in future be extended to brain-machine interfaces.