SAIL Room - 111 Levin Building, 425 S. University Avenue
Michale Fee
Department of Brain and Cognitive Sciences
McGovern Institute for Brain Research
MIT
Building a state space for song learning
Research on the avian song system has shed light on how the brain produces precise sequences that control behavior, and how the brain implements reinforcement learning (RL) of a complex behavior. While RL is a powerful strategy for learning, it depends critically on having an appropriate representation of the state space of the task. In the songbird, RL is thought to operate on a representation of song timing, but this representation is not present in young birds. I will describe a model for how the songbird brain could construct timing sequences to support RL, and will offer a hypothesis for how the auditory system could shape these sequences to align with a memory of the tutor song. This construction would also facilitate song evaluation.
A pizza lunch will be served.