Social learning through prediction error in the brain.

Area of Science:

• Cognitive Neuroscience
• Social Psychology
• Computational Psychiatry

Background:

• Social learning is crucial for survival and evolutionary fitness in social animals.
• Representing others' internal states is a long-standing interest in developmental psychology and learning studies.
• Neuroscience is investigating neural mechanisms of social learning, focusing on reward representation.

Purpose of the Study:

• To review psychological conceptualizations of representing others.
• To explore neuroscience findings on reinforcement learning signals in social learning.
• To examine neural mechanisms of social learning through reward-related information about self and other.

Main Methods:

• Review of psychological theories on representing others.
• Analysis of neuroscience studies on reinforcement learning signals.
• Discussion of self- and other-referenced reward prediction errors in brain structures.

Main Results:

• Reinforcement learning algorithms mediate social learning via reward prediction errors.
• Specific brain structures show correlates of these prediction errors for self and other.
• Computational principles underlying prediction-based learning appear conserved between self and other.

Conclusions:

• The brain utilizes similar prediction-based computational principles for processing self and other information.
• Understanding reward prediction errors offers insights into the neural basis of social learning.
• This framework highlights conserved mechanisms for navigating the social world.