Models in search of a brain

Area of Science:

• Cognitive Neuroscience
• Computational Neuroscience

Background:

• Traditional mental localization focuses on 'where' brain activity occurs, neglecting the 'what' of cognitive processes.
• Well-defined cognitive models are crucial for understanding the neural basis of cognition.

Purpose of the Study:

• To advocate for cognitive models as the primary targets for mental localization.
• To link a cognitive model of category learning to specific neural circuits.

Main Methods:

• Relating an existing category learning model to a circuit involving the hippocampus, perirhinal cortex, and prefrontal cortex.
• Simulating results from diverse participant groups (infants, amnesics, older adults) by modifying the model's response to surprising events.

Main Results:

• The model successfully simulates deficits in category learning observed in groups with varying functional integrity along the studied circuit.
• Model clusters, representing conjunctive codes, evolve from episodic roots to abstract codes through experience.
• The distinction between episodic and semantic information is shown to be fluid within the model.

Conclusions:

• Cognitive models provide a robust framework for mental localization.
• The proposed model explains dissociations between categorization and recognition based on cluster recruitment demands.
• This approach offers a unified account of memory and learning processes across different age groups and neurological conditions.