θ′ • Soil / Ukukula
Scatter
(x, y) — raw aperiodic entropy
data := {(xᵢ, yᵢ)}
θ • Roots / Ukuvula
Grounding
y(x) — encode a channel
model := y = f(x|θ)
Σ • Trunk / Ukuzula
Compression
dy/dx — invariant gradient
signal := ∂f/∂x
h(t) • Branches / Okukona
Branching
d²y/dx² — collisions/curvature
κ(x) ∝ |y''(x)|
ΔS • Canopy / Ukubona
Ledger
∫ y dx — outcomes/fruit → reseed θ′
posterior := ∫ y dx