The Simple Biological Rules That Organise the Visual Brain

Neuroscientists have long known that the brain is a marvel of efficiency, utilising 'sparse coding' to represent complex visual information using the minimum number of active neurons. However, the biological mechanisms driving this organisation remained unclear. A new study addresses this by modelling the primary visual cortex (V1) using strictly biologically grounded rules.

The researchers simulated a network of excitatory and inhibitory neurons stimulated by natural images. Crucially, they applied 'spike-timing-dependent plasticity'—where connection strengths adjust based on the precise timing of neural spikes—and homeostatic rules that regulate activity levels. Remarkable properties emerged from this substrate alone. The network spontaneously balanced itself, matching excitation with inhibition, and developed 'decorrelated' firing patterns, meaning neurons operated independently rather than shouting over one another.

These results indicate that sophisticated principles like information maximisation ('infomax') and predictive coding are not pre-installed software but emergent features. They arise naturally from the brain’s fundamental hardware, proving robust even when simulation parameters are randomised.