The geometry of abstraction
The curse of dimensionality plagues models of reinforcement learning and decision-making. The process of abstraction solves this by constructing abstract, often hidden, variables describing features shared by different specific instances, reducing dimensionality and enabling generalization in novel situations. Here we characterized neural representations in monkeys performing a task where a hidden variable described the temporal statistics of stimulus-response-outcome mappings. Abstraction was defined operationally using the generalization performance of neural decoders across novel task conditions. This type of generalization requires a particular geometric format of neural representations. Neural ensembles in dorsolateral pre-frontal cortex, anterior cingulate cortex and hippocampus, and in simulated neural networks, simultaneously represented multiple hidden and explicit variables in a format reflecting abstraction. Task events engaging cognitive operations modulated this format. These findings elucidate how the brain and artificial systems represent abstract variables, variables critical for generalization that in turn confers cognitive flexibility.
Date: 3 October 2019, 14:00 (Thursday, -1st week, Michaelmas 2019)
Venue: Sherrington Building, off Parks Road OX1 3PT
Venue Details: Large lecture theatre
Speaker: Prof Stefano Fusi (Columbia University, New York)
Organiser: Dr Tim Vogels
Topics:
Booking required?: Not required
Audience: Members of the University only
Editor: Chaitanya Chintaluri