Événements

Canonical Computation in Brains and Machines

Working memory is a cognitive process that is responsible for temporarily holding and manipulating information. Most of the empirical neuroscience research on working memory has focused on measuring sustained activity and sequential activity in prefrontal cortex (PFC) and/or parietal cortex during simple delayed-response tasks, and most of the models of working memory have been based on neural integrators. But working memory means much more than just holding a piece of information online. I will present a new theory of working memory, based on a recurrent neural circuit that I call ORGaNICs (Oscillatory Recurrent GAted Neural Integrator Circuits). ORGaNICs are a variety of LSTMs (imported from the machine learning/AI literature) that can be used to explain the complex dynamics of delay-period activity in biological neural circuits during a working memory task, and that offer some computational advantages over conventional artificial recurrent neural nets. ORGaNICs can also be applied to model motor preparation and sensory processing, commensurate with the hypothesis that executive functions, motor preparation/control, and sensory processing may share a common, canonical computational motif. Time permitting, I’ll also say a few words about inference, exploration, prediction, and the role of feedback (top-down) in the brain.
Heeger DJ, Theory of cortical function, Proc. Nat'l Acad. Sci. USA, 114:1773-1782, 2017.

Invited by the Perception, Action and Cognitive Development group

Invited by the Perception, Action and Cognitive Development group