Distributed burst firing mediates optimized cortical encoding of natural self-motion

Accurate perception of self-motion requires that vestibular input be transformed into neural representations suitable for integrating multisensory cues to guide action. Here, we demonstrate that vestibular cortical neurons represent the self-motion stimuli encountered during everyday activities in a fundamentally different manner than the artificial self-motion stimuli typically used. Using stimuli whose waveforms reproduce the head dynamics encountered in natural behavior, we found that neural activity no longer tracked stimulus velocity through graded firing-rate changes, as observed during artificial sinusoidal stimulation, but instead reliably encoded distinct motion features through burst firing. At the population level, these bursts formed a distributed code that enhanced information transmission by reducing redundancy. This transformation emerged only during natural stimulation and was absent in the vestibular thalamus. Together, our results demonstrate that the vestibular cortex constructs a distributed, feature-based representation of natural self-motion, fundamentally reshaping our understanding of how cortical circuits encode vestibular signals to mediate perception.