Dynamic mutual predictions during social learning: A computational and interbrain model

Research output: Contribution to journalReview articlepeer-review


During social interactions, we constantly learn about the thoughts, feelings, and personality traits of our interaction partners. Learning in social interactions is critical for bond formation and acquiring knowledge. Importantly, this type of learning is typically bi-directional, as both partners learn about each other simultaneously. Here we review the literature on social learning and propose a new computational and neural model characterizing mutual predictions that take place within and between interactions. According to our model, each partner in the interaction attempts to minimize the prediction error of the self and the interaction partner. In most cases, these inferential models become similar over time, thus enabling mutual understanding to develop. At the neural level, this type of social learning may be supported by interbrain plasticity, defined as a change in interbrain coupling over time in neural networks associated with social learning, among them the mentalizing network, the observation-execution system, and the hippocampus. The mutual prediction model constitutes a promising means of providing empirically verifiable accounts of how relationships develop over time.

Original languageEnglish
Article number105513
JournalNeuroscience and Biobehavioral Reviews
StatePublished - Feb 2024

Bibliographical note

Publisher Copyright:
© 2023 Elsevier Ltd


  • Empathy
  • Interbrain plasticity
  • Interbrain synchrony
  • Mutual predictions
  • Predictive coding
  • Social learning

ASJC Scopus subject areas

  • Neuropsychology and Physiological Psychology
  • Cognitive Neuroscience
  • Behavioral Neuroscience


Dive into the research topics of 'Dynamic mutual predictions during social learning: A computational and interbrain model'. Together they form a unique fingerprint.

Cite this