Demos
The demos are kept apart so that exploratory outputs cannot be mistaken for gate authority. Only the first one runs.
1. Gate evaluation and replay of this site’s determination tested
The same component as on the home page. Use step 7 to issue a linked reassessment under a later policy, an illustrative trust root, revoked evidence or a later valid time. It does not independently replay the source decision.
2. Rosetta Crosswalk: losses PROPOSED
For each domain d, define a small path category 𝖯_d. Its objects are typed states, its generating arrows are recorded transitions, and composed arrows are paths. An adapter C : 𝖯₁ → 𝖯₂ maps objects and arrows and must preserve identities and composition. It reports any object or arrow it cannot map as UNRESOLVED.
Invariants are compared through an explicit shared target J with declared maps h_i : Obj(𝖯_i) → J. For each tested state or path, the adapter checks h₂(C(s)) = h₁(s), or applies a named, versioned equivalence in J. It records witnesses and counterexamples. Calling C a functor is warranted only after identity and composition preservation have been tested over its supported scope.
No adapter has been built, so no loss table is shown. Translation never transfers standing: the target policy decides anew.
3. Sonification mapping PROPOSED
A proposed mapping from receipt states to sound, for the comparative study in paper 4. It is not implemented.
| Input | Proposed sound parameter |
|---|---|
| Check status PASS / FAIL / UNAVAILABLE / UNMEASURABLE | Timbre class (four fixed, labelled timbres) |
| Check order within a determination | Onset time, fixed interval |
| Outcome EXECUTE / HOLD | Final sustained tone vs. silence |
| Reassessment link | Repeat of the prior sequence, then the changed checks only |
Hearing a sound is not evidence of anything. A zeta-derived tone may appear only in a clearly labelled research demo, and it validates neither a phase transition nor the gate.
4. Spectral controls PROPOSED
Preregistration fields that must be fixed before any spectral run:
- Transition graph
Gconstruction and operatorL_G, with normalisation - Typed transition encoding
v_xand projection method - Observable and loss function
- Reference sequences: zeta zeros, matched random spectra (GUE-matched and Poisson), other fixed sequences
- Held-out fixtures and the criterion for added predictive or diagnostic value
- Stability checks under graph perturbation
No run has been executed. If zeta performs no better than controls, that result is published and the claim is withdrawn.