Epistemic biases in human reinforcement learning: behavioral evidence, computational characterization, normative status and possible applications

The reinforcement learning framework provides a computational and behavioral foundation for understanding how agents learn to maximize rewards and minimize punishments through interaction with their environment. This framework has been widely applied across disciplines, including artificial intelligence, animal psychology, and economics. Over the last decade, a growing body of research has shown that human reinforcement learning often deviates from normative, objective standards, exhibiting systematic biases. The aim of this paper is to propose a conceptual framework and taxonomy for evaluating computational biases within reinforcement learning. We specifically propose a distinction between praxic biases, characterized by a mismatch between internal representations and selected actions, and epistemic biases, characterized by a mismatch between past experiences and internal representations. Building on this foundation, we characterize two primary types of epistemic biases at the computational level: relative valuation and biased update. We describe their behavioral signatures, discuss their potential adaptive roles, and explore their implications for applied research. Finally, we speculate on how these findings may shape future developments in both theoretical and applied domains. Notably, despite being widely used in clinical and educational settings, reinforcement-based interventions have been comparatively neglected in the domains of public policy and decision-making, particularly when compared to more popular approaches such as nudges and boosts.