Every essay on this site eventually says the same warning in some form: the coupling field is dynamical; every fixed-K computation is a quenched approximation. This module is where that discipline stops being a footnote and becomes something you can watch. The coupling K — how strongly each pair of oscillators pulls on each other — is not a setting. It is a second dynamical system living on top of the first, with its own equation of motion, its own fixed points, its own births and deaths. When you let it run, the network stops being a machine executing on a wiring diagram and becomes something closer to a body: it reorganizes around what it is doing, and — the flagship demonstration here — it heals.
Three sections. This is an advanced module: it assumes primer 00, the learning-rule chapter (where the rule K̇ = η[R₀(K)cos(Δθ) − 2K/r] first appears), and the eigenvalues module (for the bridge finale). It complements the Lohe module's effective-coupling section: there, the wiring was chosen fresh each instant by agreement; here, the wiring learns — it has inertia, memory, and a life of its own.