Recent advances in electrophysiology and optical imaging technologies enable recordings in behaving animals from hundreds or even thousands of neurons simultaneously. Since neural coding, computation, and communication rely on coordinated activity patterns across large cell populations, such data facilitate the study of the global structure of the relationships between network activity patterns. This internal structure of neuronal population activity is an inherent attribute that cannot be revealed by analyzing functional attributes at the single-neuron level. Moreover, contemporary studies have revealed that neural activity patterns correlated with external variables undergo notable changes over the course of days and weeks—a phenomenon commonly referred to as 'representational drift.' Our research reveals that despite the observed drift in the neural code itself, the internal structure of neuronal activity remains remarkably stable. This discovery unveils constraints that exert a substantial influence on the long-term dynamics of the neural code.