Associative learning circuit based on synaptic transistors with temporal dynamics

Neuromorphic computing holds great potential for future artificial general intelligence, which requires device elements that can mimic the functionalities of biological units. Here, an ion gated synaptic transistor based on Mos₂ is demonstrated, which produces a diversity of short-term plasticity that is an enabling factor for efficient temporal information processing. A circuit based on such synaptic transistors is constructed to implement Pavlov's associative learning behavior, demonstrating the potential of the devices in brain inspired computing.