Can rodents conceive hyperbolic spaces?

Area of Science:

• Neuroscience
• Computational Neuroscience
• Cognitive Science

Background:

• Grid cells in the medial entorhinal cortex are thought to encode Euclidean space.
• Existing models propose grid cell selectivity arises from developmental self-organization, implying environmental influence.
• The potential for neural metrics to adapt beyond standard Euclidean geometry is largely unexplored.

Purpose of the Study:

• To investigate whether self-organizing models predict the formation of hyperbolic grid cells in rats.
• To explore the adaptability of neural representations of space to non-Euclidean environments.
• To propose testable predictions for experimental verification.

Main Methods:

• Utilized self-organizing models of grid cell formation.
• Simulated environments with non-Euclidean hyperbolic geometry.
• Analyzed predicted firing maps and network properties under hyperbolic conditions.

Main Results:

• Self-organizing models predict that rats raised in hyperbolic environments can form hyperbolic grid cells.
• These hyperbolic grids are predicted to exhibit multi-peaked firing maps with seven neighbors per peak, deviating from the Euclidean six.
• The formation is contingent on grid spacing relative to the hyperbolic surface's negative curvature.

Conclusions:

• Neural metrics, specifically grid cell representations, are not strictly limited to Euclidean space.
• Developmental self-organization allows for the formation of neural representations adaptable to diverse spatial geometries.
• The findings extend the concept of a universal neuronal metric to non-Euclidean planes, offering testable hypotheses.