the measure of what matters — the output of the tri-kernel when focus converges
focus is the mechanism: a conserved probability distribution over particles, $\sum \pi_i = 1$. relevance is the meaning: the judgment that emerges when that distribution reaches equilibrium. focus flows. relevance is what the flow settles on
the tri-kernel produces relevance through three complementary lenses:
- diffusion computes popularity relevance — where does probability mass accumulate through random walks? this is the PageRank intuition: a particle is relevant if many relevant particles link to it
- springs compute structural relevance — what is consistent with the graph's constraints? a particle under high tension (contradictory neighborhoods) has unstable relevance. one in a coherent cluster has robust relevance
- heat kernel computes contextual relevance — what matters at this scale? at small $\tau$, local neighborhood relevance. at large $\tau$, global thematic relevance
these three are irreducible. popularity without structure is spam. structure without exploration is echo chambers. both without scale-sensitivity miss the forest for the trees or the trees for the forest. the tri-kernel fuses all three into a single fixed point $\phi^*$ — the composite relevance of every particle in the cybergraph
cyberank is relevance materialized as a per-particle score. karma is relevance accumulated per neuron. syntropy is relevance measured as system-wide coherence. all three derive from the same $\pi^*$
the tru is the relevance machine — it reads the cybergraph and computes what matters. consensus on relevance is consensus on what matters. this is the operational definition of collective intelligence: a system that converges on relevance under conservation laws
see focus for the conserved quantity. see collective focus theorem for convergence proofs. see focus flow computation for the algorithm
discover all concepts