Brain neuromarkers predict self- and other-related mentalizing across adult, clinical, and developmental samples.

Human social interactions rely on the ability to reflect on one's own and others' internal states and traits-a psychological process known as mentalizing. Impaired or altered self- and other-related mentalizing is a hallmark of multiple psychiatric and neurodevelopmental conditions. Yet, replicable and easily testable brain markers of mentalizing have so far been lacking. Here, we apply an interpretable machine learning approach to multiple datasets (total =281) to train and validate fMRI brain signatures that predict 1) mentalizing about the self, 2) mentalizing about another person, and 3) both types of mentalizing. We test their generalizability across healthy adults, adolescents, and adults diagnosed with schizophrenia and bipolar disorder. The classifier trained across both types of mentalizing showed 98% predictive accuracy in independent validation datasets. Self-mentalizing and other-mentalizing classifiers had positive weights in anterior/medial and posterior/lateral brain areas respectively, with accuracy rates of 82% and 77% for out-of-sample prediction. Classifier patterns across cohorts revealed better self/other separation in 1) healthy adults compared to individuals with schizophrenia and 2) with increasing age in adolescence. Together, our findings reveal consistent and separable neural patterns subserving mentalizing about self and others-present at least from the age of adolescence and functionally altered in severe neuropsychiatric disorders. These mentalizing signatures hold promise as mechanistic neuromarkers to measure social-cognitive processes in different contexts and clinical conditions.