When our brains receive information via our senses, the signals are noisy and unreliable. The brain must then perform computations on this uncertain information, transforming it into an “answer” that can drive adaptive behavior.  How does the brain evaluate and represent the quality of these computations' output?  In other words, how do we know whether we've arrived at a "good" decision about stimulus identity or action selection?  In this talk, Peters will present research that aims to characterize the neural computations that underlie the ability to evaluate uncertainty, or confidence, in sensory information and in the decisions that we make based on it.  Methodologically, Peters will describe how using a combination of neuroimaging, brain stimulation, Bayesian computational modeling, electrophysiology, and behavioral approaches may help reveal how the brain computes the uncertainty in its own information processing, in order to drive adaptive behavior.  Through examining not only human neuroscience and behavior but also that of model organisms such as non-human primates and rodents, these studies aim to reveal the universal substrates of these abilities across species, as well as those aspects of uncertainty and confidence that may be uniquely human.