Expectancies in decision making, reinforcement learning, and ventral striatum

Area of Science:

• Neuroscience
• Cognitive Science
• Animal Behavior

Background:

• Decision-making involves various systems, including model-free and model-based strategies.
• Model-free strategies use prediction errors, while model-based strategies use internal forward models for evaluating outcomes.
• Expectancies in decision-making can stem from forward models or Pavlovian associations.

Purpose of the Study:

• To explore the neural basis of expectancies in decision-making.
• To investigate the role of ventral striatal neurons in reward representation during decision points.
• To connect ventral striatal activity with model-based decision strategies.

Main Methods:

• Review of neurophysiological data from van der Meer and Redish (2009a).
• Analysis of ventral striatal neuronal activity during decision-making tasks.
• Examination of hippocampal representations at decision points.

Main Results:

• Ventral striatal neurons responding to reward delivery were also activated at decision points.
• Hippocampal forward representations were observed at these same decision points.
• Co-activation suggests a potential link between ventral striatal reward signals and deliberative processes.

Conclusions:

• Ventral striatal reward representations may contribute to model-based expectancies.
• These expectancies are crucial for flexible, deliberative decision-making.
• The findings offer insights into the neural mechanisms underlying complex choices.