attractor.space
/papers/references/experiments
/notes/divergence-convergence-cycle
P001Frame-Semantic Graph Constructionframe-semanticsconstructionUpdated just now→R001TermsreferenceUpdated just now→N001Frame-type diversitydiversityframeUpdated just now→N002Coordination phasephasedynamicsUpdated just now→N003CouplingcouplingstructureUpdated just now→N004Divergence/convergence cyclecycledynamicsUpdated just now→E001ExperimentsvalidationUpdated just now→N005Fractal compositionstructureUpdated just now→N006FrameframestructureUpdated just now→N007Gini coefficientmetricUpdated just now→N008Hill diversitymetricdiversityUpdated just now→N009MembranesmembranestructureUpdated just now→R002MetricsinstrumentationUpdated just now→N010OscillatorsphasestructureUpdated just now→N011ResonanceresonanceretrievalUpdated just now→N012ScalescalestructureUpdated just now→N014SubstratestructureUpdated just now→N015Task-appropriate behaviorbehaviordiversityUpdated just now→
CC-BY-NC
lab notes on substrate dynamicsCC-BY-NC
The cycle
Failure modes
N004·Julian Fleck·Last update2026/06/25
draftv1

Divergence/convergence cycle

cycledynamics

Healthy work over the substrate cycles. Divergence opens the space — exploration, widening diversity; convergence closes it around something usable. The cycle repeats at every scale, and reading where a population sits in it is the point of watching coordination phase.

The cycle: the working set fans open (divergence) and narrows back through the constraints (convergence), over and over.

The cycle

No single phase is a failure. What counts as healthy is role-relative: a brainstorming agent should spend most of its time in exploration, a verification agent in stabilization. There is no universal healthy profile, only a match or mismatch between the substrate’s phase and the role and stakes of the task.

Failure modes

The pathologies are cycle failures. Premature convergence collapses into lock-in before exploration has done its job: the more context settles around a region, the deeper its pull, until new material cannot compete — a belief attractor. Failure to converge is drift: couplings decay without consolidation, activity without accumulation.

Premature convergence: one membrane keeps growing and siphons off every frame, until nothing competes from outside — a belief attractor.
N002Coordination phasephasedynamicsUpdated just now→