Mehrdad Jazayeri, Department of Brain and Cognitive Sciences, Investigator, McGovern Institute for Brain Research
Scalar Dynamics of Cortico-Basal Ganglia Circuits Involved in Motor Timing

Timing is central to many cognitive functions such as planning, anticipation, synchronization and mimicry. Timing also provides a unique window onto the mechanisms of internally-generated and controlled neural activity. To understand the neural code for motor timing and its relationship to neural dynamics, we analyzed single-cell and population neural activity in various nodes of the cortico-basal ganglia circuits of monkeys trained to flexibly produce different time intervals with different effectors. Single neuron had complex and heterogeneous response profiles with a notable feature that we termed scalar dynamics: response profiles were stretched or compressed in time in accordance with the desired interval. At the population-level, the scalar dynamics was substantiated by the speed with which the state of the neural signals evolved with time. Using a novel nonlinear dimensionality reduction, we established a trial-by-trial link between the animal’s produced intervals and the speed of dynamics within the subspace in which responses were temporally scaled. Analysis of the phase portrait of a recurrent neural network model of scalar dynamics revealed that the speed of the dynamics is controlled by the geometry of fixed points and their eigenmodes within the state space. These results provide a novel model for motor timing in which the produced time and its scalar fluctuations are encoded by the speed of the evolving neural states, which in turn, can be controlled flexibly by an input-dependent reorganization of the underlying attractor dynamics.

Date & Time: 
Wednesday‎, December 21‎, ‎2016‎, 9:30 - 10:30
Round ‎Table ‎Discussion‏‎: ‎11:00 ‎- ‎12:30‎‎
‎Room 221‎, ‎Dept‎. ‎Math‎. ‎Sci.‎, ‎Sharif University of Technology‎