Instructive Signals for Motor Learning

点击翻译按钮获取中文摘要

PROJECT SUMMARY Meta-learning refers to the ability to learn to learn. Whereas much progress has been made in understanding the neural mechanisms of learning, the neural mechanisms of meta-learning are a mystery. This project will investigate a candidate synaptic mechanism of meta-learning. Using integrated molecular, cellular, systems, and behavioral neuroscience approaches, the proposed research will test the hypothesis that the timing rules governing synaptic plasticity are themselves learned. Preliminary results suggest that the timing rules for associative synaptic plasticity in the cerebellum can be adaptively tuned though experience to solve what theorists have called the temporal credit assignment problem, precisely compensating for delays in the feedback about behavioral errors, so that only synapses that were active around the time that an error was generated undergo weakening during learning. If confirmed, this would represent a new dimension of the algorithms that neural circuits use to tune their own performance through experience, with broad scientific and clinical implications.