Rafael Polania, ETH Zurich
Neuro-computational and causal foundations of subjective valuation
It has been suggested that the rules guiding behavior are not arbitrary, but should follow fundamental principles of acquiring information from environmental regularities in order to make the best decisions. Moreover, these principles should incorporate strategies of information coding that take into consideration the cost of making decisions as well as biological constraints of information acquisition. In my talk, I will argue that the elaboration and understanding of such unifying theories could help to establish frameworks that allow comparing behavior not only across different tasks and goals (for instance, do I care about representing the world accurately or do I care about maximizing reward/fitness), but also across different levels of description, for instance, from single cell activity to observed behavior. In addition, I will show how it is possible to develop novel non-invasive brain stimulation interventions to reach deeper structures in the brain with the potential of selectively targeting neural circuits that compute and form subjective values. This will fundamental to directly study how experimentally altered neural activity in these brain areas causally affects value-based decisions in humans, with promising implications for neuropsychiatric conditions such as depression.