FOR 2093

C1: Translational modelling of memristive learning

The objective of project C1 is to propose circuits and evaluate concepts from other projects based on neurobiologically‐ inspired models. In particular, project C1 has to be to ascertain that those putative circuits and concepts are not only technically feasible, but also correlate to basic neurobiological mechanisms. In line with the main objectives of the research group, C1 will initially focus on the emulation of implicit learning processes and sequentially on explicit learning processes. An important task is to extract those neurobiological features, which are long-term requirements for building a cognitive circuit and which are in accordance with specific characteristics of the biological information processing system. Moreover, mandatory neurobiological functions have to be excluded from the electronic implementation. Against the background of the particular structure of the hippocampal circuit, neurobiological mechanisms will be be modeled by closely following established memory and learning mechanisms in the hippocampus of the mammalian brain. To avoid the complexity of the hippocampal structure, the developed model will conceptualize memory functions within a simplified hippocampal circuit. The goal is to achieve technically feasible reconstructions of neurobiological (hippocampal) models, which contain learning and memory functions by taking the property of the plastic storage capability of memristive devices into account.