Taking a computational approach, I develop generative models of perceptual and executive functions that closely mirror the neural circuitry underlying these functions. These models, guided by empirical observations, are lesioned in-silico to gain a mechanistic understanding of the functional outcomes of specific neural damage. This enables quantitative prediction of neural circuit responses to lesions, with the potential to identify candidate sites for targeted therapeutic intervention. I use behavioral and non-invasive neuroimaging (e.g., fMRI) data from human subjects to validate model predictions. 

My long-term research objective is to establish general principles of neural circuit reorganization and functional resilience.