Metaplasticity, a higher order of synaptic plasticity, as well as a key issue in neuroscience, is realized with artificial synapses based on a WO3 thin film, and the activity-dependent metaplastic responses of the artificial synapses, such as spike-timing-dependent plasticity, are systematically investigated. This work has significant implications in neuromorphic computation.

The synaptic weight modification depends not only on interval of the pre-/ postspike pairs according to spike-timing dependent plasticity (classical pair-STDP), but also on the timing of the preceding spike (triplet-STDP). Triplet-STDP reflects the unavoidable interaction of spike pairs in natural spike trains through the short-term suppression effect of preceding spikes. Second-order memristors with one state variable possessing short-term dynamics work in a way similar to the biological system. In this work, the suppression triplet-STDP learning rule is faithfully demonstrated by experiments and simulations using second-order memristors. Furthermore, a leaky-integrate-and-fire (LIF) neuron is simulated using a circuit constructed with second-order memristors. Taking the advantage of the LIF neuron, various neuromimetic dynamic processes, including local graded potential leaking out, postsynaptic impulse generation and backpropagation, and synaptic weight modification according to the suppression triplet-STDP rule, are realized. The realized weight-dependent pairand triplet-STDP rules are clearly in line with findings in biology. The physically realized triplet-STDP rule is powerful in developing direction and speed selectivity for complex pattern recognition and tracking tasks. These scalable artificial synapses and neurons realized in second-order memristors can intrinsically capture the neuromimetic dynamic processes; they are the promising building blocks for constructing brain-inspired computation systems.

model, and the leaky integrate-and-fire (LIF) model. The HH model, which is the closest to a biological neuron, focuses on the variation of the conductance of the ionic channels and the bioinspired spikes. However, the HH model is too complex for practical applications. [8] The functions of integration and firing in a neuron are more clearly presented in the IF and LIF models. The IF and LIF models focus on whether a neuron should fire a spike or not by comparing the local graded potential (LGP) with the threshold. The IF neuron will retain LGP boosting forever until it fires, even when it receives a subthreshold signal. Unlike the IF model, the LGP in the LIF neuron leaks out in a short time when it is lower than the threshold, which is exactly in line with what happens in a biological neuron. In addition to the leaky integrate-andfire functions, bioinspired spikes with hyperpolarization, i.e., the spike from an HH neuron, are critical in modulating firing characteristics and synaptic plasticity. For example, a hyperpolarization-activated cation current helps a neuron to respond to a synaptic change faster. [9] The transition at some synapses can also be enhanced by the hyperpolarization current. [10] Since synaptic plasticity is the basis of neuromorphic computing, fast synaptic changes will make the computing more efficient. For a neural network, especially a spike neural network, information is encoded in the spikes. Noise signals are unavoidable; however, the hyperpotential can be used as a fingerprint to remove the noise signals in the spike trains to improve the efficiency and accuracy of the neuromorphic computing.Artificial neurons based on memristive devices have been reported. An HH axon was realized with a circuit based on two Mott memristors and two parallel capacitors. [11] Simplified passive neurons consisting of parallel threshold switching (TS) and capacitors were also put forward, [12] in which accumulating charges in the capacitor increase the partial voltage of the TS device, promoting threshold switching. Afterward, the capacitor begins to discharge, forming the voltage or current pulses, i.e., spikes. Recently, Yang and co-workers constructed an artificial neuron with stochastic leaky integrate-and-fire dynamics and tunable integration time, and integrated the neuron in a fully memristive artificial neural network for pattern classification and associative learning. [13] IBM researchers implemented a stochastic neuron with IF functions based on nonvolatile phase change memristors, while the LGP was represented by the Artificial neurons with functions such as leaky integrate-and-fire (LIF) and spike output are essential for brain-inspired computation with high efficiency. However, previously implemented artificial neurons, e.g., Hodgkin-Huxley (HH) neurons, integrate-and-fire (IF) neurons, and LIF neurons, only achieve partial functionality of a biological neuron. In this work, quasi-HH neurons with leaky integrate-and-fire functions are physically demonstrated with a volatile memristive devi...

Pavlovian conditioning, a classical case of associative learning in a biological brain, is demonstrated using the Ni/Nb-SrTiO3/Ti memristive device with intrinsic forgetting properties in the framework of the asymmetric spike-timing-dependent plasticity of synapses. Three basic features of the Pavlovian conditioning, namely, acquisition, extinction and recovery, are implemented in detail. The effects of the temporal relation between conditioned and unconditioned stimuli as well as the time interval between individual training trials on the Pavlovian conditioning are investigated. The resulting change of the response strength, the number of training trials necessary for acquisition and the number of extinction trials are illustrated. This work clearly demonstrates the hardware implementation of the brain function of the associative learning.

No abstract