Abstract Social interaction is an evolutionarily conserved toolkit, critical to the survival and development of a wide variety of species. Impaired social interaction is one of the key symptoms across many neuropsychiatric disorders, including autism spectrum disorder (ASD) and schizophrenia. Therefore, elucidating the underlying neural circuits and computations of social behaviors is essential to understand the causes and mechanisms of many neurological disorders with a strong translational implication. Social interaction is dynamical in nature as it often involves a constant feedback loop of actions and reactions of all participating individuals. However, current approaches in social neuroscience often overlook this property and mostly focus on the underlying neural processes within a single individual. To fully understand how the social brain functions in health and disease, it is critical to examine the integrated system of all social participants and the neural properties that emerge from it. One of such emergent features is inter-brain synchrony. In recent years, substantial effort has been dedicated to investigating how neural dynamics across individuals are coordinated during social interaction. Using non- invasive recording techniques, many human studies have demonstrated that inter-brain synchrony emerges across social participants in various social contexts. In fact, it has also been shown that inter-brain synchrony is altered in individuals with social deficits caused by psychiatric illnesses. Despite such remarkable findings, technical constraints limit the extent of investigation and leave open various questions: how inter-brain synchrony emerges from cellular-level circuit components, and how these dynamics are related to computational processes that support healthy or impaired social interaction? Integrating a novel machine-learning approach with state-of- the-art in vivo calcium imaging in freely interacting mice, the proposed experiments will address how inter-brain synchrony (inter-brain neural correlation) arises in different genetically-defined neuronal populations in the medial prefrontal cortex (Aim 1). This work will also characterize the potential alteration of inter-brain synchrony and its behavioral implications in Shank3 mutant mice – an established ASD mouse model (Aim 2). The insights derived from this research will expand our understanding of how the information shared across multiple interacting brains can shape on-going social interaction, shedding new light onto how inter-brain synchrony can serve as a putative biomarker for impaired social interaction in ASD and laying the groundwork for new approaches to treat psychiatric illnesses.