Embedding of Low-Dimensional Sensory Dynamics in Recurrent Networks: Implications for the Geometry of Neural Representation

ArXi:2601.19019v2 Announce Type: replace-cross Neural population activity in sensory cortex is organized on low-dimensional manifolds, but why such manifolds arise and what determines their geometry remain unclear. We model cortical populations as recurrent circuits driven by low-dimensional regular sensory dynamics (circles, tori). Combining generalized synchronization and delay-embedding theory, we show that contracting recurrent networks generically develop smooth internal manifolds embedding the sensory dynamics.