Birds of a Feather: Intrinsic neural alignment and personality jointly forecast the emergence of human friendship

Human friendships are shaped by homophily, the tendency to affiliate with similar others, yet the neurobiological origins of this regularity remain unknown. Predictive-coding schemes propose that interpersonal alignment reduces uncertainty by synchronizing internal generative models. Grounded in this perspective, here we use longitudinal EEG hyperscanning to test the hypothesis that intrinsic neural similarity predicts the emergence of real-world friendships. Specifically, we recorded resting-state activity from an entire cohort of first-year university students at the beginning of the semester, when participants were strangers, and six months later, after social networks had spontaneously formed. Inter-brain synchrony and graph-theoretical global efficiency indexed dyadic neural integration, alongside Big-Five personality traits and self-reported friendship closeness. Baseline inter-brain similarity, particularly in delta and alpha bands, predicted friendship formation, and longitudinal increases in inter-brain efficiency tracked rising social closeness. Mediation analyses further showed that personality traits influenced friendship indirectly by shaping intrinsic neural alignment. Our findings demonstrate that resting-state human brain activity encodes latent interpersonal compatibility and that intrinsic neurophysiological alignment provides a neurobiological pathway linking personality to social bonding. By integrating predictive-coding theory, network neuroscience, and longitudinal social measurement, this study identifies an anticipatory mechanism through which intrinsic neural dynamics scaffold the formation of human friendships.