Abstract
The sense of agency refers to the subjective experience of controlling one's own actions and their outcomes. While agency is often thought to increase with better performance, it remains unclear how it evolves during learning. In this study, we investigated how the sense of agency changes as individuals learn when to act through reinforcement-based adaptation. We used intentional binding (IB)-a widely used, though debated, proxy measure for agency-related processes-to track temporal compression between actions and outcomes during a time-based learning task. Across four experiments, we found that IB decreased with learning, but only when feedback was imprecise yet stable, and when the outcome used to probe IB was irrelevant to the learning task. These results suggest that agency-related processes, as indexed by IB, may diminish when adaptation guides action selection, and when the outcome becomes less epistemically relevant. We discuss the possible implications of these changes in IB with learning for the sense of agency.