Abstract
To make decisions in a social context, humans try to predict the behavior of others, an ability that is thought to rely on having a model of other minds. Having such a model, known as a theory of mind, allows one to infer the intentions of others, simulate their beliefs, and predict their possible actions, taking into account the fact that others too have a similar theory of mind. Here we show that Bayesian inference of other minds explains human decisions in a group decision making task known as the Volunteer’s Dilemma. Our Bayesian model incorporates the effect of ones own actions on future rewards that could accrue to the social group one belongs to. Quantitative results from our normative model of human social decision making suggest that humans maintain a model of other minds and use this model to infer the future actions of others when deciding on their current action. We show that our model explains human data significantly better than model-free reinforcement learning and other previous models.
Footnotes
↵* Joint senior authors