SPAQUENCE

In a complex environment, animals and humans learn to navigate between sequences of significant places. Such spatial sequence learning has been shown to involve brain areas including the hippocampus (HC) and the prefrontal cortex (PFC). Recent studies show that spatial navigation in the rat hippocampus involves the replay of place cell firing during awake and sleep states generating small sequences of spatially related place cells firing. These replay episodes occur primarily during sharp-wave-ripple events. Much attention has been paid to replay during sleep in the context of long term memory consolidation. Here we focus on the role of replay during the awake state, as the animal is learning across multiple trials. We hypothesize that the generation of these short sequences of place cells during replay allow for global spatial sequence learning in PFC. We propose to develop an integrated model of the hippocampus-PFC network that is able to form spatial navigation sequences based on replay. Intellectual Merit: The proposed collaborative research will extend an existing spatial cognition model for simpler goal-oriented tasks (Barrera et al, 2011) with 1) a replay-driven model for memory formation in the hippocampus and 2) a model of spatial sequence learning in PFC. In contrast to existing work on sequence learning that relies heavily on sophisticated learning algorithms and synaptic modification rules, we propose to use an alternative computational framework known as ‘reservoir computing’ (Dominey 1995) in which large pools of prewired neural elements process information dynamically through reverberations. This reservoir will consolidate hippocampal replay sequences into larger spatial sequences that may be later recalled by subsets of the original sequences. We will constrain the model using electrophysiological recordings from the rodent hippocampal CA1 area in a multi-goal spatial task that is known to involve the hippocampus and the PFC (Watkins de Jong et al 2011). The proposed work is expected to generate a new understanding of the role of replay in memory acquisition in complex tasks such as sequence learning. That operational understanding will be leveraged and tested on robotic platforms. The originality and contribution of our proposed work includes 1) the use of hippocampal replay to create small chunks of valid trajectories, 2) the use of reservoir computing to learn spatial sequences using inputs generated by our hippocampus model, 3) our constraining and testing of the model using electrophysiological data in rats and 4) the use of the resulting model in the embodied-cognitive framework of a robot. Broader Impact: The proposal involves a multidisciplinary collaboration between a roboticist, a computational neuroscientist and an experimentalist. We plan to exploit this diversity by developing interdisciplinary material for courses and outreach activities. To actively recruit minority students, we will take advantages of REU and other programs at USF and U. of Arizona and in France. Inserm participates in multiple programs to initiate high-school students in basic research, neuroscience and robotics. We plan to explore new venues to reach the general public (especially younger students) through programs such as USF STARS and Brain Awareness Week at UA and Lyon to provide the opportunity for young students and the public to be exposed to research developments in different disciplines. In this framework, we will develop ‘sequence learning’ demonstrations involving the use of commonly found mobile robots or participants. We also propose an annual meeting in each of our 3 institutions that will be open to our respective communities and the public.