A long-lasting social attachment is built over a course of positive interactions with another individual. Forming such social bonds involves cognitive processes like perceiving salient sensory cues, learning their positive value, and selecting appropriately prosocial behavioral actions. Elucidating the mechanisms of such a complex natural behavior is a Strategic Objective of the NIMH. Our proposal will advance this Objective by contributing new knowledge within the RDoC domains of Social Processes, Positive Valence and Cognitive Systems. Our long-term goal is to enable a more vertical understanding of how molecular mechanisms influence neural circuits underlying moment-by-moment processes that must occur to build enduring social bonds. Our objective here is to use the principles of stereotypy and predictability to elucidate the behavioral and neural dynamics that underlie social bonding in the prairie vole (Microtus ochrogaster), a premier system for uncovering genetic and neuroendocrine mechanisms of social attachments. We will measure social behaviors and striatal, cortical and amygdalar neural activity over long time scales and build predictive models of the social dynamics leading to a bond. Our central hypothesis is that trajectories of stereotyped social behaviors and corresponding neural activity are predictable from a latent internal state of "pair-bondedness"; and that modeling the dynamics of this latent state will help predict future social interactions between pair-bonded prairie voles. We will quantify the predictability of such interactions in ethologically-relevant social contexts using past social behavior (Aim 1) and behavior-specific dynamic functional connectivity between those key nodes in the social brain neural network (Aim 2). Our research will have a positive impact by validating a novel, quantitative framework for studying the "dark matter" of social neuroscience, grounded in the idea that there is predictability in the dynamic processes that underlie one's trajectory through a behavioral space of stereotypical social interactions. By establishing predictability as a new paradigm for rodent social neurobiology, our studies will thus advance a comprehensive framework for how to think about social deficits and how to encourage prosocial behavior.