Dynamics of energy-efficient coding in visual cortex.

Sparse coding enables cortical populations to represent sensory inputs efficiently, yet its temporal dynamics remain poorly understood. Here, we provide direct evidence that stimulus onset initially drives broad cortical activation, transiently reducing sparseness while increasing mutual information. Over time, competitive interactions refine the population response, maintaining high mutual information as activity declines and sparseness increases. Critically, coding efficiency, quantified as the ratio of mutual information to metabolic cost, steadily improves throughout stimulus presentation, revealing an active, time-dependent optimization of sensory representations. The authors show that cortical populations refine sensory representations over time. Initially, broad activation boosts information, but as competition sharpens responses, population sparseness increases while preserving mutual information. This dynamic process steadily improves coding efficiency, revealing an active, time-dependent optimization of sensory representations that balances information and metabolic cost.