Pavlov's Dog Associative Learning Demonstrated on Synaptic-Like Organic Transistors

Abstract

In this letter, we present an original demonstration of an associative learning neural network inspired by the famous Pavlov's dogs experiment. A single nanoparticle organic memory field effect transistor (NOMFET) is used to implement each synapse. We show how the physical properties of this dynamic memristive device can be used to perform low-power write operations for the learning and implement short-term association using temporal coding and spike-timing-dependent plasticity-based learning. An electronic circuit was built to validate the proposed learning scheme with packaged devices, with good reproducibility despite the complex synaptic-like dynamic of the NOMFET in pulse regime.

Pattern classification by memristive crossbar circuits using ex situ and in situ training

“…We then present in situ training, which is more practical but also significantly more demanding because it requires adjusting the synaptic weights in parallel according to the learning rule. For the latter case, the demonstrated circuit functionality relies on the collective operation of 20 variation-prone memristors, which is at least an order of magnitude higher in complexity in comparison with that of previously reported memristor-based ANNs [23][24][25] .…”

“…The recent milestones for metal-oxide memristors, which are compatible with existing silicon technology, include demonstrations of sub-10-nm devices 16 , 10 12 -cycle endurance 17 , pico-Joule 18 and sub-ns switching 19 , and the monolithic integration of several memristive crossbar layers 20 . These milestones, in turn, revived interest in the development of memristor-based ANNs and led to numerous few-or single-memristor demonstrations of synaptic functionality and simple associative memory [21][22][23][24][25][26][27][28] , as well as the theoretical modelling of large-scale networks 10,[29][30][31][32][33] . Despite significant progress in memristor crossbar memories 20,[34][35][36][37] , memristor-based ANNs have proven to be significantly more challenging and have yet to be demonstrated.…”

Low voltage and time constant organic synapse-transistor

“…A compact model was developed 9 and we demonstrated an associative memory, which can be trained to exhibit a Pavlovian response. 10 However, these devices were limited to work with spikes in the range of few tens of volts and time scale of 1-100s. Both fields (neuro-inspired computing and bioelectronics) require devices working at lower voltages (to save energy consumption during computing and because action potentials in synapse and neurons have amplitudes of around 100 mV) and higher speed (e.g.…”

Reconfigurable Threshold Logic Gates using memristive devices

“…While several papers have proposed using memristors [12][13][14][15][16], most are just simulation [17], often without a circuit [18]. A few papers go beyond simulation to design circuits, at least for a portion of the system [19,20], and even fewer design the complete circuit and build a working prototype [21][22][23][24]. Papers on physical memristors usually evaluate them in isolation using a signal analyzer [25,26].…”