Neural indicator of feedback sensitivity characterizes reinforcement learning behavior in adults

Reinforcement learning (RL) uses feedback evaluation that helps individuals learn how to adjust strategies based on the outcomes of their actions. This study explored systematic differences in a neural indicator of feedback sensitivity, measured by feedback related negativity (FRN), and its impact on RL behavior. Participants were separated into two groups based on the FRN-based neural indicator. The high feedback sensitivity (HFS) group, characterized by greater differentiation in FRN between feedback types (reward vs. punishment), exhibited poorer behavioral accuracy and reduced spatial targeting. In contrast, individuals with low feedback sensitivity (LFS), showing smaller neural differences across feedback types, demonstrated improved learning performance and spatial accuracy. These findings suggest that overreactive responses to feedback in the HFS group may disrupt the integration of information needed for learning, leading to poorer outcomes. Interestingly, a similar pattern to that of FRN and frontal theta power was observed in P300 amplitude, highlighting the potential role of reward salience and attention allocation in feedback-based learning. Overall, the findings indicate that reduced feedback sensitivity, as indexed by FRN, appears to support learning, providing insights into how adaptive learning can be enhanced.