A parietal grid-like code rotates with cognitive maps but lags rapid behavioral transfer.

Area of Science:

• Neuroscience
• Cognitive Science
• Computational Neuroscience

Background:

• The neural grid code is theorized to enable relational knowledge generalization and transfer for behavioral adaptation.
• Understanding how grid representations change during context transitions and their link to behavior is limited.

Purpose of the Study:

• To investigate if neural grid representations rotate to align with behavioral goals during context transitions.
• To determine if these rotations underpin knowledge transfer and rapid behavioral adaptation.

Main Methods:

• fMRI was used to measure neural representations in human participants.
• Participants performed a task involving abrupt, synchronous rotations of target locations at unsignaled state changes.
• Behavioral performance and neural activity were analyzed in relation to context transitions.

Main Results:

• Participants demonstrated rapid inference of target locations, including novel ones, after state changes, indicating zero-shot transfer.
• A cognitive grid-like code was observed in the right posterior parietal cortex, with its phase angle rotating with target positions.
• The observed rotation of the parietal grid code was too slow to explain rapid behavioral improvements; these were linked to frontoparietal and orbitofrontal cortex activity.

Conclusions:

• Humans exhibit rapid knowledge transfer capabilities, adapting behavior in novel circumstances.
• A parietal grid-like code does rotate into behaviorally relevant frames, but its role in rapid adaptation is questionable.
• Alternative neural mechanisms beyond parietal grid code rotations likely support swift knowledge transfer and behavioral adaptation.