Action and outcome encoding in the primate caudate nucleus.

Area of Science:

• Neuroscience
• Cognitive Science
• Computational Neuroscience

Background:

• The basal ganglia are implicated in reinforcement learning.
• Previous research focused on pre-movement activity for action selection.
• Post-movement activity in striatal neurons suggests a role in evaluating action outcomes.

Purpose of the Study:

• To investigate the role of post-movement striatal neurons in reinforcement learning.
• To determine if individual neurons encode saccade direction, reward reception, or both.
• To understand the functional segregation of post-movement neuronal activity in the caudate nucleus.

Main Methods:

• Recorded activity from phasically active neurons in the caudate nucleus of monkeys.
• Utilized a probabilistically rewarded delayed saccade task.
• Analyzed neuronal responses in relation to saccade execution and reward outcome.

Main Results:

• A majority of recorded neurons (77/149) showed peak activity after saccade execution.
• Many post-movement neurons were tuned to the direction of the preceding saccade (61/77).
• Separate neuronal subsets encoded saccade direction (43/60) or reward outcome (35/60) during the reward epoch.
• Little overlap was observed between neurons encoding action and outcome.

Conclusions:

• Striatal neurons active after movement are segregated into distinct populations.
• One population encodes action (saccade direction), while another encodes outcome (reward).
• These separate populations provide complementary information for reinforcement learning.