Researchers show that sending a single, well‑tuned driving signal to one neuron can stop extreme events across several model neuronal networks. They use a drive–response setup and test the idea on three network shapes: a pair of coupled neurons, a single-layer network of many neurons, and a two-layer (multiplex) network. Each network is made from FitzHugh–Nagumo neuronal units, a standard mathematical model of neuron activity, and the unforced response networks produce extreme events on their own.

In the experiments the team influences just one neuron in the response network with a dominant driving signal. They find that this single influence is enough to reduce or remove the extreme events in every network configuration they try. The work is presented as a validation across the three distinct topologies rather than a proof for all possible networks.

The mechanisms behind control differ by topology. In the two‑neuron case, suppression happens when the drive breaks the phase‑locking between the driving neuron and the targeted response neuron — in other words, the timing of their oscillations no longer lines up in the way that had produced extremes. In the single‑layer and multiplex networks, control works by disrupting what the authors call protoevent frequency dynamics and by causing the driven neuron to decouple in frequency from the rest of the network. Put simply, the driven neuron stops sharing the rhythmic patterns that were triggering the large events.

The authors also report a scaling advantage: when the number of response neurons connected to the drive is increased, the onset of control occurs earlier. That is, connecting the drive to a larger group of neurons makes the suppression effect appear sooner, suggesting the method may become more effective as the driven neighborhood grows.