Cortical circuits exhibit variable yet bounded activity patterns, suggesting operation near—but not within—fully chaotic regimes. Here we develop a minimal three-variable rate model for primary visual cortex (V1) that reveals how biologically motivated feedback mechanisms can function as intrinsic chaos controllers. We adopt the simplest known chaotic Lotka–Volterra system as a phenomenological scaffold and introduce three biologically motivated modifications: excitatory-to-inhibitory (E \(\rightarrow \) I) feedback coupling, homeostatic regulation of modulatory drive, and orientation-tuned sensory input. These modifications transform the excitatory (E), inhibitory (I), and modulatory (M) population dynamics from chaotic strange attractors into controlled limit cycles—a 93% reduction in dynamical variance. The model reproduces key V1 phenomena: orientation selectivity matching experimental distributions (OSI \(= 0.38 \pm 0.09\) ), stimulus-induced variability quenching, and realistic spiking irregularity when coupled to Hodgkin–Huxley neurons (CV \(_{\text {ISI}} = 0.27\) , within in vitro range). Parameter space analysis reveals that feedback mechanisms robustly stabilize activity across most of the tested chaotic regime. We further demonstrate that, within this minimal structure, the specific \((1-\textrm{I}^2)\) disinhibition nonlinearity enables chaos—bounded alternatives tested do not support chaotic dynamics. Our findings suggest that cortical circuits possess an intrinsic capacity for chaos that is actively suppressed by canonical feedback motifs, positioning the brain at the edge of instability where computational flexibility meets reliable signal processing.
