Cognitive computational modeler interested in flexible decision making, cognitive control, and the prefrontal cortex.
Currently I am working as a PhD student in Belgium in the lab of Clay Holroyd. Our main focus is understanding the human capability for the execution of extended sequences and the (partial) generalization of actions between these sequences. This work combines Hierarchical Reinforcement Learning (HRL) theory with both fMRI and electrophysiological data. We aim to show that the dorsal anterior cingulate cortex (dACC), also called the midcingulate cortex (MCC), holds distributed representations of task and task-set variables which can speed up and disambiguate decisions. We use recurrent neural networks to model these distributed representations, which should emerge after training the network on several related tasks (meta-learning). We then use representational similarity analysis (RSA) to compare these representations to neural activity during the same tasks.