Abstract:
A hallmark of human intelligence is our ability to quickly adapt to new situations (“tasks”) in an uncertain world. Many such tasks involve a learning process, i.e. adjusting ongoing behavior in light of experiences within the task. Adapting to a new task is thus “meta-learning”, i.e. producing a learning strategy tailored for it. This is thought to be solved by the brain by building fast internal representations for task states in the prefrontal cortex (PFC), yet with largely unknown mechanisms. Building on my previous work on working memory and task learning, my goal is to study 1) how task-state spaces are implemented in the PFC when a learning strategy for a task is already formed ; 2) how task-state spaces are formed in the PFC when a learning strategy is created for a new task ; 3) the role of working memory in doing so.
About the speaker: Dr Flora Bouchacourt
Dr Flora Bouchacourt is a postdoc in computational neuroscience at Princeton University, in the labs of Dr Tim Buschman and Dr Nathaniel Daw. She studies the representation and dynamics of working memory as well as meta-learning of tasks. She received a BSc and MSc in theoretical physics from Ecole Polytechnique, Paris, France. She did her Ph.D. in computational neuroscience in the lab of Dr Srdjan Ostojic at Ecole Normale Superieure.
Click here to register via Eventbrite.