A03Cognition and sociality Neurocomputational primitives for decision-making with use of models of others

 We must make decisions in our social environment, taking account of social factors and consideration for others. A fundamental of this human cognitive ability comprises two basic neurocompuational primitives in brain: modeling others’ hidden and internal process to predict their behavior, and using the predictions for making appropriate and thus better decisions. In this research, we address the two primitives and thereby aim to investigate the underlying neural mechanisms and computations for the use of others’ models in making social decisions. We are also interested in how the primitives may be modified and modulated if their revision is necessary, for instance, switching or re-learning. For these purposes, we use a solid computational framework such as reinforcement learning and extend it to social situations and combine such modeling with a novel experimental paradigm in human fMRI experiments in our investigation.