Read Montague - Virginia Tech Carilion School of Medicine The nascent field of Computational Psychiatry seeks to re-cast mental illness in terms of derangements in computations that underlie human cognition. This is a tall order. Human cognition is a multi-scale phenomenon with important supporting mechanisms that operate on small spatial and temporal scales up to interactions among groups of humans. In addition to this multi-scale character, there is also the problem that human neurobiology has not previously had methods available to provide the kinds of transformative insights that have been flowing out of modern work on model organisms. This latter work has provided some unique biological insights into cellular and circuit-level phenomena that support behavior, but there is still the difficult jump (‘the software problem’) to understanding how model organism behavior relates to human cognition. Human behavior can be driven by a range of behavioral phenomena not easily mimicked in model organisms and so it’s paramount to pursue direct, rigorous approaches using humans. In this talk, I will introduce our own approach to invasive investigation of neuromodulatory systems (dopamine and serotonin) in conscious human brain and review the genre of models used to capture the computations thought to be carried in part by these systems.