Neurocomputational modeling of rule abstraction and memorization during probabilistic stimulus-reward learning

Area of Science:

• Cognitive Neuroscience
• Decision Science
• Neuroeconomics

Background:

• Preferential choice relies on learning strategies like rule abstraction or memorization.
• Rule abstraction is used for systematic stimuli, while memorization handles unsystematic ones.
• Understanding the neurocognitive mechanisms underlying these strategies is crucial for decision-making research.

Purpose of the Study:

• To investigate how humans deploy rule abstraction and memorization for object selection.
• To identify the distinct learning trajectories and neural correlates of these two strategies.
• To reveal the neurocognitive mechanisms differentiating rule abstraction and memorization.

Main Methods:

• Functional magnetic resonance imaging (fMRI) to measure brain activity.
• Eye-tracking to monitor visual attention and fixation patterns.
• Cognitive modeling to capture learning trajectories and strategy deployment.

Main Results:

• Differential learning trajectories and fixation patterns were observed, corresponding to rule discovery and incremental learning.
• Cognitive modeling successfully captured these distinct learning processes.
• Overlapping brain networks for cognitive control and value-based decision-making were identified.

Conclusions:

• Multi-modal data and model-informed analyses linked specific brain regions to rule abstraction and memorization.
• The study reveals the neurocognitive mechanisms differentiating these two fundamental learning strategies.
• Findings advance our understanding of decision-making and learning in humans.