Project Summary Our long-term goal is to elucidate the fundamental neurobiological mechanisms supporting social learning. Many of our everyday social decisions require constant assessments of other individuals, such as whether they can be trusted. These decisions are inherently risky, as it is often uncertain, especially with strangers, how outcomes will unfold. Despite this, the mechanisms governing social risky learning remain largely unexplored. A hypersensitivity to risk and uncertainty—a hallmark symptom of anxiety that often results in a pronounced and maladaptive bias toward making risk-avoidant choices—provides an ideal test bed to probe the mechanisms governing social risky learning. The objective of this project is to innovatively merge methodologies and insights from associative learning models and neuroecononmics to examine the functional properties of the brain-behavior relationships that mediate social learning under uncertainty, while also identifying how alterations in these learning mechanisms shift socially risky behavior in maladaptive ways. Our central hypothesis is that a social learning model can capture the neurobiological mechanisms governing both healthy and maladaptive social risk taking. Our specific aims will 1) discover how social value (e.g. trustworthiness) is behaviorally and neurally instantiated in uncertain environments, 2) determine the role of affect in biasing these social learning processes, and 3) uncover knowledge about the relationship between anxiety and social learning and how it can lead to maladaptive socially risky choices. By providing a computational account of this relationship, we may show that social risky avoidant behavior emerges at the level of value assignment learning. Such a finding would highlight that individuals avoid socially risky choices because of a failure in affective learning. This contribution is significant since it will elucidate both optimal behavioral patterns and dysfunction and pathology during social learning—findings that may reveal potential biomarkers to aid in diagnosis and targeted interventions in those suffering from anxiety. Finally, the proposed research is innovative because it harnesses emerging computational, neuroscience, and theoretical knowledge on nonsocial learning in order to develop a deeper understanding of social risk-taking and its link with anxiety. !