syntropy

the order in the cybergraph, measured in bits. syntropy is the quantity tru exists to grow — the purpose of the whole pipeline. it is the negentropy of the focus distribution: how sharply collective attention concentrates on real structure rather than spreading uniformly across noise.

$$J(\phi^*) = \sum_j \phi^*(j)\,\log\big(|V|\cdot\phi^*(j)\big) = D_{\mathrm{KL}}\big(\phi^* \,\|\, u\big)$$

where $\phi^*$ is the tri-kernel fixed point, $u$ is the uniform distribution over $|V|$ particles, and $D_{\mathrm{KL}}$ is the KL divergence. a flat $\phi^* = u$ gives $J = 0$ — maximum entropy, zero order, pure noise. a $\phi^*$ that sharpens around a connected, useful core gives large $J$ — high order. syntropy is exactly the distance the graph has travelled from noise toward structure.

why it is the purpose

intelligence is the capacity to compress experience into structure that predicts. syntropy measures that structure directly. a graph with high syntropy has found the patterns that matter; a graph with low syntropy is still dominated by spam, redundancy, and noise. maximizing syntropy is maximizing the graph's predictive grip on its domain — it is the key metabolic factor of superintelligence.

tru recomputes syntropy every epoch alongside $\phi^*$ — both emitted by focusing. it is the network's vital sign: high and rising means the collective is learning; falling means noise is winning.

relation to free energy and reward

syntropy is the macroscopic face of the free energy the tri-kernel minimizes. the same iteration that drives $\phi^*$ to the variational free-energy minimum (see tri-kernel §2.4) drives syntropy up. minimizing surprise and maximizing order are one motion seen from two sides.

at the level of a single neuron, syntropy is the BTS score $s_i$: the bits of information one neuron's cyberlinks added to the collective picture. a neuron whose links sharpen $\phi^*$ contributes positive syntropy and earns; a neuron whose links add noise contributes negative syntropy. the proven per-signal shift is the impulse $\Delta\phi^*$, and rewards mint against it. syntropy growth is therefore the thing the economy pays for — the link between order and value is direct.

syntropy is the task-free measure of collective order. its task-grounded sibling is superadditivity $\sigma$ — the advantage of the collective $\phi^*$ over any individual neuron's ego focus on link-prediction and retrieval. $J$ asks whether the focus is structured; $\sigma$ asks whether that structure beats every part. they are distinct axes: benchmark (superadditivity) shows $\sigma$ rises with algebraic connectivity $\lambda_2$ while $J$ falls with it — densifying a graph spreads focus toward uniform, so connectivity buys collective advantage but costs syntropy.

see focus for the distribution syntropy measures · superadditivity for the collective-intelligence measurement · Bayesian Truth Serum for the per-neuron version · rewards for how syntropy growth becomes $CYB · tri-kernel for the convergence that produces $\phi^*$.

discover all concepts