<p>The floating vegetation hydrodynamically divides the flow into an upper “in-canopy sluggish zone” with reduced momentum and a lower “bilaterally confined advective zone” with higher momentum. Accurately defining the vertical extent of the in-canopy sluggish zone (described by the penetration depth) is critical for assessing the ecological functions of floating vegetation. In this study, a genetic programming (GP) algorithm was employed to establish an explicit predictive model quantifying this boundary. The model indicates that the penetration depth is governed by the Froude number, the vegetation slenderness ratio, the relative vegetation height, and the blockage ratio. Aside from achieving high predictive accuracy, the derived model reveals the competitive mechanisms controlling the development of the in-canopy sluggish zone. The zone’s extent is not a monotonic function of single parameters but rather the outcome of the dynamic balance between shear-driven instability and drag-induced damping. Compared with existing models, the proposed model demonstrates significant advantages in both applicability and accuracy, thus providing an effective tool for the assessment of aquatic organisms’ habitats.</p>