Neurotransmitter-Mediated Plasticity in 2D Perovskite Memristor for Reinforcement Learning

影响因子：19.0

DOI码：10.1021/acsami.3c12674

发表刊物：Advanced Functional Material

关键字：2D perovskites;memristors;neurotransmitter dopamine;reinforcement learning;synaptic plasticity

摘要：The neuromorphic computing architecture is a promising artificial intelligence for implementing hierarchical processing, in-memory computing, event-driven operation and functional specialization in computing systems. However, current investigations mainly focus on unisensory processing without objective experience which is contrary to the flexible sensory learning capability in the human brain that can sense and process information according to the ever-changing environment. For example, a dominant paradigm for reconfigurable bio-learning features is the emotional experience. The neurotransmitter dopamine is released during arousal, influencing the vital brain functions involved in cognition, reward learning, movement and motivation. Here, the on-demand configuration of a biorealistic synaptic connection based on a 2D CaTa2O7 (CTO) device is demonstrated that can be adaptively reconfigured for a reinforcement learning purpose by the light-active resistive switching, which originated from the photon-regulated metaplasticity. The low energy consumption of 12.4 fJ endows the reinforcement learning system with high power efficiency and reliability. Finally, in-sensor computing with a CTO synapse is implemented with a filtering function to process digital data in a neuromorphic engineering manner. This work demonstrates the feasibility of 2D perovskite neuromorphic device with enhanced biological plausibility in the approaching post-Moore era.

合写作者：Shidong Chen, Xiaohan Cheng, Wang Chen, Ziyu Xiong, Ziyu Lv, Chunyan Wu, Li Wang, Guohua Zhang, Xiaobo Zhu, Linbao Luo, Su-Ting Han

第一作者：Yan Wang

通讯作者：Lin-Bao Luo, Su-Ting Han

文献类型：Article

卷号：34

期号：3

页面范围：2309807(1-9)

ISSN号：1616-301X

是否译文：否

发表时间：2023-10-12