Binocular Rivalry: Competing Neural Populations

When each eye is shown a different image, perception does not blend the two — it alternates, spontaneously and unpredictably, between one image and the other every few seconds. This simulator renders the leading computational account of that switch: two rate-coded neural populations, one per eye, locked in mutual inhibition and slowed by spike-frequency adaptation, with independent noise driving irregular timing. Watch two clusters of model neurons compete in real time, tune the inhibition, adaptation and noise that drive the fight, and bias which eye's image wins more often — the same dynamics thought to underlie real binocular rivalry and, more broadly, how competing neural representations resolve into a single conscious percept.