Definition
Neural Grouping
A neural grouping is a set of neurons linked by synapses that mutually excite and stimulate one another, so that the group’s output is determined by the combination of inputs it receives, not by any single neuron in isolation.
Formally, let be a subgraph of the brain’s neural graph, with neurons and excitatory synapses . Given an input pattern over , the group produces an output pattern determined by recurrent excitation within :
where is the strength of the synapse from to and is the neuronal activation function. The mapping is many-to-many: different input combinations yield different output combinations, even within the same group.
Mechanism
The mechanism rests on three properties:
- Mutual excitation. Neurons in the group reinforce one another’s activity through recurrent synapses, so activation propagates and persists within the group.
- Combinatorial input–output. Output depends on which subset of inputs is co-active and on their relative strengths, not on any single input channel alone.
- Intensity sensitivity. A weak input may drive some members of the group above threshold while leaving others silent; a stronger input recruits more members, producing a distinct output pattern.
The same group produces two distinct outputs depending on input strength: a weak signal to recruits but leaves below threshold, while a strong signal drives both above threshold.
Origin
Alexander Bain introduced neural groupings in Mind and body (1873) as the physical realisation of associationist ideas. Neurons that fire together wire together — Bain’s statement of what later became Hebbian learning — is the learning rule that shapes which neurons fall into the same grouping. The grouping, not the individual neuron, is the unit that carries the mental representation.