Prediction mismatch responses arise as corrections of a predictive spiking code

Abstract

Prediction mismatch responses in cortex seem to signal the difference between an internal model of the animal and sensory observations. Often these responses are interpreted as evidence for the existence of error neurons, which guide inference in models of hierarchical predictive coding. Here we show that prediction mismatch responses also arise naturally in a spiking encoding of sensory signals, where spikes predict the future signal. In this model, the predictive representation has to be corrected when a mispredicted stimulus appears, which requires additional neural activity. This adaptive correction could explain why mismatch response latency can vary with mismatch detection difficulty, as the network gathers sensory evidence before committing to a correction. Prediction mismatch responses thus might not reflect the computation of errors per se, but rather the reorganization of the neural code when new information is incorporated.