Memory-reliant Post-error Slowing Is Associated with Successful Learning and Fronto-occipital Activity

Area of Science:

• Cognitive Neuroscience
• Neuroimaging
• Reinforcement Learning

Background:

• Negative feedback in cognitive tasks influences future action devaluation and response caution.
• Reinforcement learning and cognitive control theories explain these phenomena separately.
• The interaction between cognitive control and value updating under uncertainty remains unclear.

Purpose of the Study:

• To investigate cognitive control-based behavioral adjustments during probabilistic reinforcement learning.
• To examine the influence of these adjustments on performance in a subsequent test phase.
• To explore the neural correlates of these adaptive behaviors using fMRI.

Main Methods:

• Functional magnetic resonance imaging (fMRI) was employed during a probabilistic reinforcement learning task.
• Behavioral adjustments, specifically post-error slowing, were analyzed.
• The relationship between behavioral adjustments and brain activity (BOLD signal) was assessed.

Main Results:

• Post-error slowing during reinforcement learning was associated with improved performance in a later test phase.
• Adjusting response speed after negative feedback correlated with brain activity in the right inferior frontal gyrus and bilateral middle occipital cortex.
• Activity in the bilateral middle occipital cortex showed partial overlap with unsigned prediction error signals.

Conclusions:

• Functionally relevant, memory-reliant behavioral adjustments like post-error slowing are crucial for learning.
• Cognitive control and feature processing regions interact to facilitate feedback-congruent adaptations.
• These findings suggest a mechanism for adaptive learning under uncertainty involving neural interactions.