Recording electrode
Neuron soma + axon
⚠ Couldn't load the 3D engineThree.js failed to load from the CDN. Check your connection and reload.
Every brain-computer interface that reads "what did the brain intend" starts from a much messier signal: a single implanted electrode sitting near several neurons at once, each contributing its own overlapping action potential to one shared voltage trace. This simulation renders that recording site in 3D — an electrode tip surrounded by two to four neurons at different distances — and streams the resulting raw waveform to a live oscilloscope trace. A real spike-sorting pipeline runs alongside it: each detected spike is measured for peak amplitude and width, plotted in a 2D feature space, and classified back to its source neuron by nearest-centroid assignment. Turning up recording noise or adding more simultaneous neurons pushes the feature clusters together and visibly drags sorting accuracy down, which is exactly the trade-off real multi-electrode neural interfaces have to manage before any higher-level decoding can happen.