Neural commonalities and dissociations of human social and experiential learning

Humans navigate the world by learning from both social interactions and direct experiences. Although these two learning strategies are essential for adaptive survival, a systematic neural comparison between them has been lacking. Here, we combined quantitative meta‑analysis with large‑scale network mapping to identify the shared and distinct brain systems underlying social and experiential learning (as represented by Pavlovian conditioning) in healthy humans. Both learning modes engaged common regions involved in value computation, such as the ventral striatum and anterior insula. However, they showed largely dissociable network patterns across the brain: social learning was primarily linked to networks involved in social cognition, whereas experiential learning was predominantly associated with reward and cognitive control. These distinct connectivity profiles reliably differentiated the two learning modes at both aggregate and individual levels. Additionally, we found that appetitive and aversive forms of social learning were supported by separate brain networks. Taken together, our findings provide convergent evidence for how the human brain flexibly reuses core value-processing circuits while engaging specialized networks tailored to distinct learning demands.