abstract Anita Tusche

Anita Tusche (Max Planck Institute for Human Cognitive and Brain Sciences, Leipzig)

Neurocomputational Approaches to Social and Non-Social Decision-Making

How do people flexibly represent varying values and goals? Does the pursuit of distinct self-control goals recruit similar mechanisms? Are some people generally better at self-control, or does it depend on the domain? Using an innovative combination of computational modeling and multivariate decoding of fMRI data, we reveal a common neurocomputational mechanism in the dorsolateral prefrontal cortex (DLPFC) that supports current goals for dietary and social behavior via context-sensitive representations of choice-relevant attributes. These computations in DLPFC also predicted an individuals’ regulatory success across choice domains. However, I show that flexible representations of social attributes such as others’ welfare were encoded only in the temporoparietal junction (TPJ) and precuneus, revealing distinct self-regulatory mechanisms as a function of choice domain. My results identify important divisions of labor in neural computations supporting value and self-control, and explain when and why the capacity for self-control generalizes (or doesn’t) across contexts and choice domains.