Learning shapes neural geometry in the primate prefrontal cortex

AI Summary

PFC representations evolve during learning from high-dimensional, nonlinear, randomly mixed codes to low-dimensional, rule-selective formats.
When generalising to new stimuli, PFC representations become abstract and stimulus-invariant, supporting rule transfer.
These results reconcile conflicting views by showing how PFC geometry adapts across learning stages, balancing exploration and generalisation.

Nat Neurosci. 2026 Jun 25. doi: 10.1038/s41593-026-02333-w. Online ahead of print.

ABSTRACT

The relationship between the geometry of neural representations and the task being performed is a central question in neuroscience. The primate prefrontal cortex (PFC) is a primary focus of inquiry, as it can encode information with geometries that either rely on past experience or are experience agnostic. One hypothesis is that PFC representations should evolve with learning, from a format that supports exploration of all possible task rules to a format that minimizes the encoding of task-irrelevant features and supports generalization. Here we test this idea by recording neural activity from the macaque PFC when learning a new rule (‘XOR rule’) from scratch. We show that PFC representations progress from being high dimensional, nonlinear and randomly mixed to low dimensional and rule selective. Upon generalizing the rule to new stimuli, these representations further evolve into an abstract, stimulus-invariant geometry. These findings reconcile previously conflicting accounts of PFC function by demonstrating how neural representations adapt across distinct stages of learning.

PMID:42350815 | DOI:10.1038/s41593-026-02333-w

Document this CPD