1. What can one anchored arrow see?
Hold A, B and C fixed; select the source's exact proper hinge of D about AB. The incident faces' intrinsic geometry and the complete local AB operator row stay fixed. ABD's ambient normal and the dihedral angle do change.
The universal remote-sensitivity conjecture was falsified: the symmetric AB packet is shielded even though the full resolvent changes. Unequal incident weights are an explicit constitutive control, not a discovered source law. Six field coordinates do not erase the twelve labelled arrow occurrences.
Experiment and exact shielding argument · Full response receipt
2. Two different electron states; one identical link reading
The declared candidate link is electron raw face0 spin0 → proton raw AB:x, weight 1. The source provides two different logical doublets with equal norm and identical first reading. These are alternatives, not two occupied electrons.
The displayed amplitudes are the receipt's unnormalized logical inputs. The first link has a one-complex-dimensional dark direction. Two permitted complex readings reconstruct the doublet; retaining/reproducing input, applying a rotation, and accessing amplitudes are mathematical permissions—not derived laboratory capabilities. The persistent unit relation is not a complete spinor decoder.
The repaired potential attachment retains all 24 fold interiors with position and mode separate. This is a different information-loss problem; retaining those witnesses alone does not repair the link kernel.
Link and fold-retaining attachment · Exact pair, observations and inverses · Why this is not yet physical binding
3. Same 1,836 transfers; different information tasks
| Task | Protonlike | Neutronlike |
|---|
The common aggregate staged spectrum does not determine the stage-resolved task. Read the selected coordinate addresses at raw, local and aligned copies; no extra raw Gx reading is inserted.
The rank is evaluated from the source's all-subset theorem, not estimated from a few examples. Coordinates and metrics differ between the carriers. Neither transfer counts nor reconstruction thresholds are particle masses.
Neutron tasks and metric-aware comparison · Neutron receipt · Proton all-mask theorem and decoders
4. A reference has a resource and a lifetime
For a scalar Clifford-unit input, the existing conservative exchange stores it, delays it, then returns it. The product-preserving Cl(2,0) representation sends the unit to I. The pinned comparison reference requires an explicitly prepared two-unit resource and explicit boundary identifications.
Pinned SOFTWARE transport: 137 → 32 → 32
Inherited target keys preserve and reflect exactly 977 original leaf-pair contexts. This explorer reproduces the source fixture's endpoint enumeration. It does not identify family endpoints with assembly A→B or introduce connectors.
Original target fibres and their retained members
| Original target | Leaves retained through both promotions |
|---|
What retaining the branch contrast adds
The outer branch observer retains branch masses and one matrix contrast in addition to unchanged J4, so it can recover both branch means without reopening histories. It is not a physical detector or J4 alone.
Stored and delayed operators can be equal while their history/depth differs. The active bank is empty after the third exchange; history is not a free permanent physical reference. Preparation/reset buttons create new declared mathematical inputs, not cost-free physical reuse. A cross-target pair is rejected: reachability, an available frame matrix or equal values do not authorize a connector.
End-to-end conditional binding · Memory protocol · Port binding · Pinned software receipt
5. An internal phase clock—but only locally
The existing frozen-contact Hamiltonian creates cross-topology coherence from the actual incoming mixture. The reading χ=Tr(ρQ) can parameterize the state only where its derivative is nonzero.
All forty source-coordinate populations at this reading
Only the three audited χ samples are selectable; no invented interpolation or new integration is shown. Selecting 3H displays the receipt's verified rate-independent state at the same χ. Disabled matching leaves the input stationary and rejects the clock chart. These populations are phase-record occupations, not physical formation probabilities. The bare gate J²=I does not accumulate this phase; this is not physical seconds, a global clock, or a derived arrow-processing law.
Construction and clock boundary · Full densities, operators and residuals
6. Conservation accuracy is not spatial convergence
Four fixtures compare one coherent packet at K=3.30 ε and RMS width 0.035 ε, with matched omitted-mode energy—not matched mode counts. The baseline is supplied, not freshly rerun as a full packet.
| Fixture | Strain / radial grid | Modes | First omitted energy / ε | Selected observable |
|---|
Numerical audits—separate from spatial accuracy
| Fixture | Full-S tests | Prepared-column tests | Equation residual | Raw audit residual |
|---|
Expansion-directed transport survives these finite refinements; the combined peak magnitude changes by about −27.482%. This is an observed grid difference, not an error bound. Each new fixture has one full-S central-energy test and 32 prepared-column tests, not 33 full-S tests. Tilt/hinge grids remain fixed, closed modes remain truncated, and only one packet center is tested. No continuum certificate, physical species or calibrated speed law is inferred.
Full methods and limitations · Archive's verified numerical summary