
NeuroNexus · neuronexus.com · Glossary
Unit (Single-Unit and Multi-Unit Activity)
A unit is a neuron inferred from an extracellular recording: a cluster of spikes whose waveform is consistent enough to be attributed to one cell. Single-unit activity is the spikes of one isolated neuron; multi-unit activity is the summed spikes of several neurons near a site that cannot be told apart.
The distinction decides what a recording can support. Single units allow tuning curves, response latencies and spike timing per cell; multi-unit activity reports how active the local population is, without sorting, and tolerates drift.
The electrode sets which is possible. Within about 50 µm of a site a neuron's spike is large and well formed and can be isolated; between 50 and about 100 µm several neurons overlap at similar amplitudes and are recorded as multi-unit activity; a layout that puts several sites inside one detection radius — a tetrode, a polytrode, a high-density shank — lets a spike sorter separate cells that a single site would merge. A unit is called single when its spike amplitude is at least about three times the noise, fewer than about 1% of its inter-spike intervals violate the refractory period, and its waveform is stable over the session. Multi-unit activity is usually quantified as threshold crossings or spike-band power on a site. NeuroNexus designs for unit isolation on this basis: the tetrode, polytrode and Buzsaki layouts exist to put several sites inside one radius, and 177 µm² iridium sites are the working standard because they isolate units at an impedance, 200 kΩ to 1 MΩ at 1 kHz, that the headstage amplifier handles without added noise.
Typical values
| Single-unit isolation, typical criteria | amplitude ≥ 3 × noise; < 1% refractory violations; stable waveform |
|---|---|
| Multi-unit activity | spikes of about 2 to 10 neurons within about 50 to 100 µm of a site |
| Refractory period | about 1 to 2 ms |
Related Terms
Cite this term
NeuroNexus. "Unit (Single-Unit and Multi-Unit Activity)." NeuroNexus Glossary, updated 2026-10-12. https://www.neuronexus.com/resources/glossary/unit