Decision-Making in Repeated Games: Insights from Active Inference

This review systematically explores the potential of the active inference framework in illuminating the cognitive mechanisms of decision-making within repeated games. Characterized by multi-round interactions and social uncertainty, repeated games more closely resemble real-world social scenarios, where the decision-making process involves interconnected cognitive components such as inference, policy selection, and learning. Unlike traditional reinforcement learning models, active inference, grounded in the free energy minimization principle, unifies perception, learning, planning, and action within a single generative model. Belief updating is achieved by minimizing variational free energy, while the exploration-exploitation dilemma is balanced by minimizing expected free energy. Formulated based on partially observable Markov decision processes, the framework naturally incorporates social uncertainty, and its hierarchical structure allows for simulating mentalizing processes, thereby offering a unified account of social decision-making. Future research can further validate its effectiveness through model simulation and behavioral fitting.