Calendar

Abstract: 

The brain is a complex system comprised of billions of neurons that work coherently together to control our behavior and general function. Large scale recordings of neural activity over broad anatomical areas with high spatial and temporal resolution are increasingly common in modern experimental neuroscience, generating datasets manifesting rich and varied dynamics. To tackle the scale and complexity of the observed dynamics, numerous approaches based on a dynamical systems framework have been proposed. Here we use AI/ML approaches to capture multi-scale dynamics and recapitulate cognitive processes at the neuron level. We also consider how neural properties, at the individual neuron and population level, explain adaptation during learning. To this end, we examine learning during a neurofeedback task with perturbations. We find that certain key properties are predictive of how critical a neuron or population is to learning and how well the animal can adapt their neural activity within the task framework. 

PRESENTER INFORMATION