ACC representations of reward-driven motivation over hierarchically-organized behavior

Area of Science:

• Neuroscience
• Cognitive Science
• Computational Neuroscience

Background:

• Reward-guided behavior is extensively studied, but its effect on hierarchical sequential behavior production is unclear.
• The anterior cingulate cortex (ACC) is hypothesized to motivate hierarchical task completion by encoding goals in neural ensembles.
• Understanding ACC's role in reward modulation of these representations is crucial.

Purpose of the Study:

• To investigate how rewards modulate neural representations in the ACC during hierarchical sequence production.
• To apply model-based representational similarity analysis (RSA) to an adapted recurrent neural network (RNN) model of ACC activity.
• To compare model predictions with fMRI-BOLD responses during a hierarchical sequence production task.

Main Methods:

• Model-based representational similarity analysis (RSA).
• Adaptation of a recurrent neural network (RNN) model of anterior cingulate cortex (ACC) activity.
• fMRI-BOLD data acquisition during a hierarchical sequence production task.

Main Results:

• A cluster in the ACC showed high similarity to the model's reward modulation predictions.
• ACC activity discriminated rewarded from non-rewarded sequences primarily at the first trial step.
• Reward-related ACC activity positively correlated with task performance (accuracy and response time).

Conclusions:

• The ACC monitors reward contingencies to motivate control over hierarchically-organized, goal-directed behaviors.
• ACC's role in encoding reward prediction errors at the onset of action sequences is supported.
• Findings highlight ACC's unique contribution to reward-modulated sequential behavior.