demeurons

Demeurons are a theoretical class of neurons defined by their membership in locally interacting neural populations called demes. The term combines deme with neuron, reflecting a modular organization proposed in some models of neural computation.

In this framework, a deme is a discrete neural population with high intra-deme synaptic connectivity and relatively weaker, sparser connections to neurons in neighboring demes. Demeurons are the principal cells that participate in these local circuits, contributing to modular processing, localized feature representation, and pattern separation within a deme.

Developmentally, demeuron networks are assumed to arise from activity-dependent clustering and selective synaptic strengthening within deme boundaries, with plasticity mechanisms promoting both stability within a deme and occasional cross-deme integration for global coherence.

Evidence for demeuron organization is primarily theoretical and indirect, drawn from analyses that identify modular structure in cortical activity and from computational models showing improved learning in modular networks. Critics note that the term is not widely used in empirical neuroscience and that it risks over-generalization when applied to real brain tissue.

In practice, demeurons are used in discussions of modularity and small-world organization, and in neuromorphic and artificial intelligence research as a framework for designing segmented, locally coherent networks. See also modular neural networks, cortical columns, and neural microcircuits.