Dynamic and stable population coding of attentional instructions coexist in the prefrontal cortex

Area of Science:

• Neuroscience
• Cognitive Neuroscience

Background:

• Neural representations in the prefrontal cortex (PFC) dynamically adapt to task demands.
• Understanding how these dynamic coding schemes depend on encoded variables and anatomical constraints is crucial.

Purpose of the Study:

• Investigate the anatomically specific neural coding of spatial and color attentional instructions in the PFC.
• Examine the dynamics of neural representations and identify mechanisms for stable information retention.

Main Methods:

• Utilized a cued attention task combined with multivariate classification methods.
• Analyzed neuronal ensemble activity in the frontal eye field (FEF) and ventrolateral PFC (vlPFC).
• Applied high-dimensional and low-dimensional analyses to neural activity.

Main Results:

• Spatial and color attentional instructions were decoded from PFC neuronal ensembles in an anatomically specific manner.
• Frontal eye field (FEF) showed robust spatial decoding, while ventrolateral PFC (vlPFC) excelled in color decoding.
• vlPFC exhibited dynamic coding for color, yet stable, time-invariant color information was found within a low-dimensional subspace.
• Spatial attention modulated feature decoding in vlPFC more profoundly than in visual area V4.

Conclusions:

• Dynamic population coding in PFC is shaped by anatomical constraints.
• Stable subspace coding allows for time-invariant decoding of attentional information, coexisting with dynamic network states.