# Dimensional Reality: Court-Grade Mathematical and Physical Audit

**Version:** 0.7  
**Date:** 2026-09-15  
**Status:** Adversarial research memorandum, not a claim of experimental confirmation

## Executive verdict

The framework has a mathematically valid core:

1. An observation can be modeled as a projection from a richer state space.
2. Hidden distinctions matter exactly when states that look identical now have different future observable laws.
3. Adding enough hidden state can make a projected description dynamically closed.
4. Wrapped phase loses orientation and winding information; quadrature restores orientation, while an integer winding register preserves history.
5. Quantum interference is reduced when path information becomes encoded in any physical degree of freedom.
6. Ordinary linear single-particle quantum mechanics has no irreducible interference of order three or higher; this gives an eight-coordinate alignment hypothesis a sharp, though difficult, eight-path null test.
7. Eight phase coordinates have an exact, non-metaphorical realization as the eight computational-basis phases of a three-qubit equal-magnitude state; removal of global phase leaves exactly seven observable Walsh/Pauli interaction coordinates.
8. Individual-outcome prediction now has an exact frontier: a precommitted accessible lift must earn positive held-out log evidence against the complete Born conditional distribution; a sequential likelihood-ratio e-process supplies an anytime-valid test.
9. The frontier now remains valid under a declared composite Born null, charges every compiler-generated lift a prefix-code search penalty, and inherits the projection-collision discipline developed in RRLS.
10. For a Born-marginal-preserving extension, the optimal evidence rate of an accessible lift is exactly its conditional mutual information with the outcome; a new three-channel audit now separates that information from marginal Born violation, future-setting leakage, remote signaling, and ordinary apparatus memory.
11. P3-R1 now instantiates the single-device outcome frontier as an unactivated IBM Quantum-compatible pilot: predictions for both settings precede delayed setting generation, three prefix-coded eight-cycle probes preserve the ideal marginal by construction, calibration and ledgers are hash-bound, and an independent replay agrees on adversarial synthetic controls.

Those statements are defensible. They do **not** yet prove eight physical dimensions, a soul, consciousness-triggered collapse, cosmological expansion by observation, or a unique new law of nature. Several central mathematical ideas also have close prior art in quotient dynamics, Markov lumpability, computational mechanics, and predictive-state representations.

The strongest responsible conclusion after reconstructing closed-loop, delayed-choice, non-Markovian, macroscopic-superposition, and multipath experiments is therefore:

> **Dimensional Reality is presently a coherent mathematical research framework with an eight-cycle physical hypothesis—not a proven physical theory.**

No empirical theory can honestly be made “unrefutable.” A theory protected from every possible observation is scientifically weaker, not stronger. The court-grade target is instead: explicit assumptions, correct theorems, clear prior art, a unique quantitative prediction, and a result that could prove the proposed physical law wrong.

---

## I. Vocabulary that must not be mixed

| Term | Precise use in this framework | What it does not establish |
|---|---|---|
| Coordinate | A number used to label a model state | A physical dimension |
| Latent variable | A state component not included in the current observation | Extra spacetime |
| Dimension | Number of independent coordinates locally required by a specified state space | Ontological reality by itself |
| Phase | A cyclic coordinate, usually modulo \(2\pi\) | A new dimension automatically |
| Winding | Integer count of completed phase cycles in a chosen lift | Physical observability without a coupling law |
| Observation | A physical record or data-generating map | Human consciousness specifically |
| Decoherence | Loss of locally accessible interference through entanglement with other degrees of freedom | A universally agreed solution to the measurement problem |
| Zero | An origin, null value, or initial possibility specification | “All potential” unless a possibility space and measure are supplied |
| \(G\) | A proposed metaphysical ground or soul variable | A physical variable until an operational rule connects it to data |

The distinction between a mathematical coordinate and a physical dimension is non-negotiable. Writing eight numbers produces an eight-coordinate model. It does not, without empirical consequences, show that nature has eight dimensions.

---

## II. Frozen mathematical core

### 1. Eight-coordinate state and alignment

Let the underlying state be

\[
z=(z_1,\ldots,z_8)\in\Omega\subseteq\mathbb R^8.
\]

Define the normalized diagonal direction

\[
\mathbf e=\frac{1}{\sqrt 8}(1,1,1,1,1,1,1,1).
\]

The **perfect-alignment line** is

\[
\mathfrak Z_8=\{a\mathbf e:a\in\mathbb R\}.
\]

The unique vector-space zero is

\[
\mathbf O_8=(0,\ldots,0).
\]

It is important to call \(\mathfrak Z_8\) the alignment line, not “the zero” itself. Every point on the line has equal coordinates, but only \(\mathbf O_8\) is the additive identity.

Let

\[
P_{\parallel}=\mathbf e\mathbf e^{\mathsf T},\qquad
P_{\perp}=I-P_{\parallel}.
\]

An exact alignment defect is

\[
\Delta_8(z)=\lVert P_{\perp}z\rVert.
\]

Then

\[
\Delta_8(z)=0\iff z\in\mathfrak Z_8.
\]

This is a theorem by construction, not evidence that nature uses the construction.

### 2. “All potential” requires a possibility object

A zero vector contains eight zeros; it does not encode a family of alternatives. “All potential” can instead be represented by a possibility space and initial measure:

\[
\mathsf Z_0=(\Omega,\Sigma,\mu_0),
\]

where \(\Omega\) is the set of admissible states, \(\Sigma\) is a collection of measurable events, and \(\mu_0\) assigns initial weights. If no probabilities are intended, replace \(\mu_0\) by an admissible set \(\mathcal P_0\subseteq\Omega\).

This makes the idea precise:

- \(\mathbf O_8\): coordinate origin;
- \(\mathfrak Z_8\): equal-coordinate alignment;
- \(\mathcal P_0\) or \(\mu_0\): potential possibilities.

They are different mathematical objects and should remain different.

### 3. Cyclic time: wheel, helix, and recurrence

For cycle \(i\), define unwrapped phase, wrapped phase, and winding by

\[
\Phi_i(t)=\omega_i t+\phi_i,
\qquad
\theta_i(t)=\Phi_i(t)\bmod 2\pi,
\qquad
n_i(t)=\left\lfloor\frac{\Phi_i(t)}{2\pi}\right\rfloor.
\]

The wrapped state \(\theta_i\in S^1\) is the wheel. The pair \((\theta_i,n_i)\) is a lifted helix. Eight wrapped phases live on the torus \(\mathbb T^8\).

#### Exact-recurrence theorem

There is a common recurrence time \(T>0\) satisfying

\[
\theta_i(t+T)=\theta_i(t)\quad\text{for every }i\text{ and }t
\]

if and only if

\[
\omega_iT\in2\pi\mathbb Z\quad\text{for all }i.
\]

For nonzero frequencies, this is equivalent to rational commensurability of all frequency ratios \(\omega_i/\omega_j\).

**Consequence.** “Time can align only at the beginning” is not a universal theorem. It is true of exact phase alignment after \(t=0\) when the relevant frequencies satisfy the necessary rational-independence condition. Even then, torus flows can return arbitrarily close to an earlier alignment, so exact nonrecurrence must not be confused with practical nonrecurrence.

### 4. The four seasons

The four seasons can be represented as four sectors of an annual phase \(\theta_y\in S^1\), but equal quarter-sectors are an idealization. Astronomical seasons are anchored to solstices and equinoxes, and meteorological/ecological seasons vary by convention and location. The defensible statement is:

\[
\text{season}=S(\theta_y,\text{latitude},\text{climate},\ldots),
\]

not “four seasons prove four time dimensions.”

---

## III. Observation as projection: the strongest theorem

Let \(T:\Omega\to\Omega\) be the underlying deterministic evolution and let

\[
\pi:\Omega\to Y
\]

be the observation map. The observer records \(y=\pi(z)\), not necessarily the full \(z\).

### Deterministic closure theorem

There exists a well-defined observed evolution \(L:Y\to Y\) satisfying

\[
\pi\circ T=L\circ\pi
\]

on \(\pi(\Omega)\) if and only if

\[
\pi(x)=\pi(x')\Longrightarrow \pi(Tx)=\pi(Tx').
\]

#### Proof

If \(\pi T=L\pi\) and \(\pi(x)=\pi(x')\), then

\[
\pi(Tx)=L(\pi x)=L(\pi x')=\pi(Tx').
\]

Conversely, suppose equal present observations always give equal next observations. For \(y\in\pi(\Omega)\), choose any \(x\in\pi^{-1}(y)\) and define

\[
L(y)=\pi(Tx).
\]

The stated condition makes this definition independent of the chosen representative \(x\). Therefore \(L\) is well-defined and \(\pi T=L\pi\). ∎

### Meaning

The fiber

\[
\pi^{-1}(y)=\{x\in\Omega:\pi(x)=y\}
\]

contains all full states that look identical to the observer now. A hidden distinction is predictively relevant exactly when two points in the same fiber evolve to different future observations.

This is the precise, defensible form of the intuition that “we see only a projection.”

### Probabilistic closure and lumpability

Let \(K(x,\cdot)\) be a Markov transition kernel. A closed observed kernel \(Q(y,\cdot)\) exists when

\[
\pi_*K(x,\cdot)=Q(\pi(x),\cdot),
\]

where \(\pi_*\) denotes the pushforward to observable outcomes. In finite or countable state spaces, this holds if and only if

\[
\pi(x)=\pi(x')
\Longrightarrow
K(x,\pi^{-1}(B))=K(x',\pi^{-1}(B))
\]

for every observable set \(B\subseteq Y\). This is a form of strong lumpability. Necessary and sufficient lumpability criteria are an established area of Markov-chain theory; see [Jacobi and Görnerup (2008)](https://arxiv.org/abs/0710.1986).

Define the total-variation closure defect

\[
\mathcal D(\pi)=
\sup_{\pi(x)=\pi(x')}
d_{\mathrm{TV}}\!\left(\pi_*K(x,\cdot),\pi_*K(x',\cdot)\right).
\]

Then

\[
\mathcal D(\pi)=0
\]

exactly when the observed dynamics is closed. If \(\mathcal D(\pi)>0\), any single proposed transition law attached to that observed state must disagree with at least one of the two hidden-state laws by at least \(\mathcal D(\pi)/2\), by the triangle inequality.

This quantity is experimentally useful: it measures how badly the chosen observable state fails to be predictively sufficient.

### Minimal predictive completion

For histories \(h,h'\), define

\[
h\sim_{\mathrm{pred}}h'
\iff
\Pr(Y_{\mathrm{future}}\mid h,\mathrm{do}(u))
=
\Pr(Y_{\mathrm{future}}\mid h',\mathrm{do}(u))
\]

for every allowed future intervention \(u\). The equivalence class \([h]_{\mathrm{pred}}\) is a predictively sufficient state.

This construction is powerful, but not wholly new. Computational mechanics reconstructs minimal predictive dynamics from measurement sequences [Crutchfield and Young (1989)](https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.63.105), and predictive-state representations encode state through future action-conditional observations [Littman and Sutton (2001)](https://proceedings.neurips.cc/paper/2001/hash/1e4d36177d71bbb3558e43af9577d70e-Abstract.html).

**Novelty verdict:** the projection/fiber language is a valid foundation and may support a novel synthesis, but it cannot honestly be claimed as an unprecedented theorem without a sharper difference from these literatures.

---

## IV. Oriented lifted phase: what is known and what could be new

Suppose a sensor reports

\[
y_t=\cos\Phi_t.
\]

The same observation is produced by \(\Phi=+\theta\) and \(\Phi=-\theta\). One step later,

\[
\cos(\theta+\omega)-\cos(-\theta+\omega)
=-2\sin\theta\sin\omega,
\]

which is generically nonzero. Therefore \(y_t\) alone is not a closed state.

Add the quadrature

\[
q_t=\sin\Phi_t.
\]

Then the pair obeys the closed rotation

\[
\begin{pmatrix}y_{t+1}\\q_{t+1}\end{pmatrix}
=
\begin{pmatrix}
\cos\omega&-\sin\omega\\
\sin\omega&\cos\omega
\end{pmatrix}
\begin{pmatrix}y_t\\q_t\end{pmatrix}.
\]

The proposed **oriented lifted phase address** is

\[
\Lambda_t=(q_t,n_t).
\]

The verdict is split:

- \(q_t\) repairs a real predictive ambiguity, but phase quadratures are standard mathematics and physics.
- \(n_t\) stores how many times the cycle has wrapped, but it is physically redundant whenever all laws depend only on \(e^{i\Phi_t}\).
- Winding becomes a new physical variable only if two states with the same complete standard physical state but different \(n_t\) produce different future outcome distributions.

For eight phases, a generic cosine-only observation has sign ambiguity in every coordinate plus arbitrary winding:

\[
\pi^{-1}(y)\sim\{\pm1\}^8\times\mathbb Z^8.
\]

However, if a common phase shift

\[
(\theta_1,\ldots,\theta_8)\mapsto
(\theta_1+\alpha,\ldots,\theta_8+\alpha)
\]

is a gauge symmetry, only seven relative phases are independent:

\[
\mathbb T^8/S^1\cong\mathbb T^7.
\]

An eighth phase is operational only if an external reference breaks that common-shift symmetry.

### Physical-newness criterion

Let \(Z_{\mathrm{std}}\) contain the full standard state and all accepted control variables. The proposed hidden address is physically new only if

\[
I(Y_{\mathrm{future}};\Lambda_8\mid Z_{\mathrm{std}})>0
\]

under a preregistered intervention, or equivalently if some matched states satisfy

\[
\Pr(Y_{\mathrm{future}}\mid Z_{\mathrm{std}},\Lambda_A)
\ne
\Pr(Y_{\mathrm{future}}\mid Z_{\mathrm{std}},\Lambda_B).
\]

Without that inequality, \(\Lambda_8\) is a useful bookkeeping coordinate, not a discovered physical variable.

---

## V. Quantum audit

### 1. Double slit and which-path information

For balanced paths with a physical path marker,

\[
|\Psi\rangle=
\frac{1}{\sqrt2}
\left(
|L\rangle|D_L\rangle
+e^{i\phi}|R\rangle|D_R\rangle
\right).
\]

At screen position \(x\), the interference term is multiplied by

\[
\gamma=\langle D_L|D_R\rangle.
\]

If the marker states are identical, \(|\gamma|=1\) and full coherence is possible. If they are orthogonal, \(\gamma=0\) and the local interference term disappears. In the ideal balanced pure-state case,

\[
V=|\gamma|,
\qquad
D=\sqrt{1-|\gamma|^2},
\qquad
D^2+V^2=1,
\]

with the more general inequality \(D^2+V^2\le1\). Englert derived the visibility/which-way bound [in 1996](https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.77.2154), and Dürr, Nonn, and Rempe independently measured distinguishability and visibility in an atom interferometer in agreement with the duality relation [in 1998](https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.81.5705).

If eight independent environmental channels factorize,

\[
|D_L\rangle=\bigotimes_{k=1}^8|h_k^L\rangle,
\qquad
|D_R\rangle=\bigotimes_{k=1}^8|h_k^R\rangle,
\]

then

\[
\gamma_8=\prod_{k=1}^8\langle h_k^L|h_k^R\rangle.
\]

This gives the framework a clean geometric picture: every channel can reduce the overlap. But the product rule follows from ordinary tensor-product quantum mechanics. Calling the eight channels “dimensions” does not turn it into new evidence for eight spacetime dimensions.

### 2. Schrödinger’s cat and decoherence

A simplified system-environment state is

\[
|\Psi\rangle=
a|\mathrm{alive}\rangle|E_a\rangle
+b|\mathrm{dead}\rangle|E_d\rangle.
\]

After tracing out the environment, the cat’s reduced density matrix has off-diagonal terms multiplied by

\[
\Gamma=\langle E_d|E_a\rangle.
\]

As the environment records distinguish the branches, \(|\Gamma|\) becomes small and local interference becomes inaccessible. Mesoscopic “cat” states and their progressive decoherence were directly observed by Brune and collaborators [in 1996](https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.77.4887).

This supports the physical-record explanation. It does not prove that consciousness selects an outcome, and the experiment requires no human awareness at the moment the environmental record forms. Decoherence explains suppression of observable interference; interpretations still differ on what, if anything, selects one experienced outcome.

### 3. What “observer” can safely mean

For empirical physics, define an observation as a stable physical record:

\[
\text{observation event}
=
\text{system-record correlation that can be read later}.
\]

A person may later experience that record, but substituting “soul awareness” for the record-forming interaction adds a metaphysical postulate. That postulate is not automatically false; it is simply not yet an experimentally identified cause.

### 4. Bell constraint

A local hidden-variable model ordinarily assumes, among other conditions,

\[
P(a,b\mid x,y,\lambda)
=P(a\mid x,\lambda)P(b\mid y,\lambda)
\]

and measurement independence of \(\lambda\) from the freely chosen settings. Loophole-free Bell experiments reject the relevant local-realist null. Hensen and collaborators reported \(S=2.42\pm0.20\) against the CHSH local bound \(S\le2\), with an event-ready experiment over 1.3 km [in 2015](https://www.nature.com/articles/nature15759).

Therefore an eight-dimensional hidden-state model cannot merely say “the missing coordinates locally predetermine both outcomes.” It must explicitly be at least one of the following:

- nonlocal while remaining operationally no-signaling;
- contextual;
- non-deterministic at the hidden level;
- rejecting measurement independence with a clear mechanism;
- or outside the tested assumptions in another precisely stated way.

Merely saying “higher dimensions connect everything” does not perform this work.

### 5. Contextuality constraint

Noncontextual hidden variables assign a measurement’s value independently of which compatible measurements accompany it. A trapped-ion experiment found a state-independent violation of a noncontextual bound, reporting values from \(5.23(5)\) to \(5.46(4)\) where the relevant noncontextual limit is \(4\) [Kirchmair et al. (2009)](https://arxiv.org/abs/0904.1655).

Therefore any outcome-assignment rule for \(G\), the soul, or hidden phase must be contextual—or must explain exactly why the tested operational assumptions do not apply.

### 6. Quantum-state reality constraint

The Pusey–Barrett–Rudolph theorem shows, under its preparation-independence assumption, that models treating distinct pure quantum states as merely overlapping information about the same underlying physical state contradict quantum predictions [Pusey, Barrett, and Rudolph (2012)](https://arxiv.org/abs/1111.3328).

This does not eliminate all deeper ontologies. It does mean that a “zero contains all possibilities and the wavefunction is only ignorance” story must specify whether preparation independence holds and how the ontic states reproduce quantum statistics.

### 7. No-signaling constraint

Quantum nonlocal correlations do not permit controllable faster-than-light communication. Under standard state and probability assumptions, imposing no superluminal signaling strongly constrains dynamics to linear completely positive maps on density matrices [Simon, Bužek, and Gisin (2001)](https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.87.170405). Finite-speed hidden-influence explanations of all quantum correlations can themselves lead to operational superluminal signaling in multipartite scenarios [Bancal et al. (2012)](https://arxiv.org/abs/1110.3795).

Any soul-selection or higher-dimensional coupling must therefore satisfy, for spacelike separated parties,

\[
\sum_bP(a,b\mid x,y)=P(a\mid x)
\]

independently of the remote choice \(y\), and symmetrically for the other party.

### 8. Multi-observer constraint

Extended Wigner’s-friend experiments test combinations of assumptions about observer-independent facts, locality, and freedom of choice. A six-photon experiment violated an associated Bell-type inequality by five standard deviations; the authors’ conclusion is explicitly conditional on retaining locality and free choice [Proietti et al. (2019)](https://arxiv.org/abs/1902.05080).

This does **not** prove that consciousness creates reality. It means a participatory-reality model must define how multiple observers’ records combine without contradiction and state which assumption in the no-go construction it rejects.

---

## VI. The record-fiber version of participatory reality

Let \(\Omega\) be a set of candidate complete histories, let the random variable \(W\in\Omega\) denote the realized complete history, and let \(R_t(W)\) be all physical records accessible by time \(t\). The worlds compatible with the current record are

\[
\mathcal F_t(r)=\{\omega\in\Omega:R_t(\omega)=r\}.
\]

If later records preserve earlier ones, then for the realized history,

\[
\mathcal F_{t+1}\subseteq\mathcal F_t.
\]

For these random variables on a probability space,

\[
H(W\mid R_{t+1})\le H(W\mid R_t)
\]

in expectation. Thus two directions coexist:

- the **domain of recorded experience** can grow;
- the **set of histories compatible with the accumulated record** can shrink.

This is a rigorous rendering of the Minecraft analogy. It is epistemic/record-theoretic expansion. It is **not yet** the cosmological expansion of the metric, whose quantitative observable is a scale factor \(a(t)\) or equivalent geometric dynamics.

To equate experiential expansion with cosmic expansion, the theory would need a bridge equation such as

\[
\frac{\dot a}{a}=F(R_t,\mu_t,G,\ldots),
\]

derive a specific \(F\), and outperform existing cosmological models. No such derivation currently exists.

---

## VII. Status of \(G\) and the soul hypothesis

The clean metaphysical formulation is:

1. \(G\) is a ground of being not represented as an ordinary spacetime coordinate.
2. A person’s soul supplies a first-person center of experience.
3. Physical reality supplies a branching or probabilistic structure of possible records.
4. Experience follows one consistent record path.

As philosophy or theology, this is coherent enough to discuss. As physics, the crucial probability law is missing:

\[
P(o\mid M,\psi,G)=\;?
\]

There are only two empirical possibilities:

- If the right side always equals the Born rule \(\langle\psi|E_o|\psi\rangle\), \(G\) is empirically silent in those experiments.
- If it differs, the theory must state the difference before observing the data and must preserve Bell correlations, contextuality, no-signaling, and agreement among observers.

The existence of first-person experience is a datum. The inference that an entity outside spacetime selects quantum outcomes is an additional premise, not a mathematical consequence of projection, phase, or decoherence.

---

## VIII. Claim ledger

| Claim | Status | Court-grade verdict |
|---|---|---|
| Equal observed states yield a closed observed dynamics iff their next observed states agree | Proven within the model | Exact theorem |
| Probabilistic closure is equality of pushed-forward transition laws on each observation fiber | Proven in the stated setting | Established lumpability structure |
| Total-variation closure defect vanishes iff the projection is predictively closed | Proven by definition and the closure criterion | Useful diagnostic |
| Cosine-only phase loses orientation; sine quadrature closes fixed rotations | Proven | Standard phase-space mathematics |
| Exact common recurrence requires commensurate frequencies | Proven | Corrects the unconditional nonrecurrence claim |
| Eight factorized environmental overlaps multiply | Proven under factorization | Standard tensor-product quantum mechanics |
| Born probabilities have no irreducible single-particle interference above second order | Proven under linear, context-independent amplitudes | Supplies an exact eight-path null test |
| Eight equal-magnitude three-qubit phases reduce to seven physical relative phases | Proven after global-phase quotient | Exact realization of \(\mathbb T^8/S^1\cong\mathbb T^7\) |
| The eight phases invertibly map to one global, three local, three pair, and one triple Pauli-\(Z\) coefficient | Proven for a fixed unwrapped lift | Standard Walsh/diagonal-unitary structure |
| Computational-basis probabilities alone can have closure defect one | Proven with a common future Hadamard intervention | Relative phase is maximally predictively relevant |
| No pre-outcome predictor beats a complete exact Born conditional law in expected log score | Proven from nonnegative KL divergence | Born Prediction Wall theorem |
| Coordinatewise composite-null envelopes produce a valid outcome e-process | Proven for predictable null families satisfying \(p_t(y)\le u_t(y)\) | Conservative protection against calibration uncertainty |
| Prefix-coded mixtures protect a frozen countable lift search | Proven from Kraft's inequality and convex e-process mixing | Each description bit carries an evidence penalty |
| In a Born-marginal-preserving extension, a candidate decoder's log-evidence rate equals accessible conditional information minus its conditional KL misspecification | Proven | Exact Information-Rate identity; a nonpreserving extension has an additional marginal Born-divergence term |
| A finite tested context set determines an unrestricted law on untested contexts | False | Explicit extension collision agrees on seven contexts and differs by total variation (0.8) on the eighth |
| A compressed lift with an unequal-outcome-law collision can be repaired by a stronger downstream decoder | False | Information erased by the projection is unavailable to every downstream decoder |
| An accessible variable predicting real Born-random outcomes has been found | Unsupported | Synthetic positive worlds test the detector only |
| The phase lift alone constitutes a universal quantum computer | False | Noncommuting gates, measurement, error correction, scaling, and hardware are additionally required |
| More physical which-path information reduces interference | Established theory and experiment | Strong evidence; not unique to this framework |
| Environmental decoherence occurs without conscious observation | Established physical mechanism and experiment | Undermines the necessity of consciousness for fringe loss |
| Accumulating records shrink the compatible-history fiber | Proven given record nesting | Epistemic result, not cosmological expansion |
| Nature has exactly eight physical dimensions | Unsupported hypothesis | No derived necessity or unique observation yet |
| Eight raw phases imply eight physical degrees of freedom | Generally false without an external phase reference | Common phase may be gauge; seven relatives remain |
| Winding is physically observable | Open hypothesis | Requires history-matched future divergence |
| A soul selects quantum outcomes | Metaphysical hypothesis | No operational law or evidence presently supplied |
| Human observation expands physical space | Unsupported physical identification | Record-domain growth is not metric expansion |
| Magnetic repulsion is an invisible material wall | False literally | It is electromagnetic interaction, not hidden matter filling the gap |
| The framework solves the measurement problem | Not established | It redescribes records and decoherence but lacks a unique outcome law |
| The current synthesis is wholly unprecedented | Not supported | Significant prior art exists; specific synthesis may still be original |

### Internal computational evidence

Earlier finite and synthetic checks can support implementation correctness but not universal physics:

- 7,662 finite closure/refinement checks passed in the internal test suite.
- The previously proposed “seven strict refinements” theorem was retracted because the closure operator is idempotent: after exact closure, repeating the same operator produces no further strict refinement.
- Additional synthetic suites over rational worlds, recurrence sequences, damping configurations, and excluded predictions test code and definitions only.
- Synthetic cycle tomography does not identify arbitrary hidden coordinates without a restricted response model.

The number seven may be a chosen register size or the number of relative phases after quotienting a common phase. It is not, from these checks, a theorem that physical reality has seven hidden dimensions plus one observed dimension.

---

## IX. Prior-art audit

| Framework component | Closest established structure | Remaining possible contribution |
|---|---|---|
| Projection \(\pi\) and closed observed evolution | Quotient dynamics / semiconjugacy | Unified notation and application across perception and quantum records |
| Probabilistic fibers | Strong lumpability | Closure-defect diagnostic tied to experiments |
| Predictively equivalent histories | Computational mechanics / causal states | A physically specified eight-cycle restriction, if derived |
| Action-conditional predictive state | Predictive-state representations | Connection to dimensional phase addresses |
| \((\cos\Phi,\sin\Phi)\) | Standard quadrature/phase-space representation | None by itself |
| Winding register \(n\) | Universal covering space / topological winding | New only if a measurable law depends on absolute winding |
| Product of hidden-channel overlaps | Standard decoherence/tensor products | New only if “eight” and the channel couplings are independently fixed |
| Record-compatible world fibers | Bayesian conditioning / filtration | A careful metaphysical interpretation, not new physical evidence |

The court-safe novelty claim is:

> “We propose a synthesis that uses observation fibers, lifted cyclic state, and record consistency to search for a specific winding-sensitive physical law.”

The unsafe claim is:

> “We have already proved an eight-dimensional theory of reality.”

---

## X. Minimal falsifiable bridge model

The existing axioms do not imply any new measured number. To make a first empirical target, introduce an explicitly provisional bridge axiom.

### Bridge axiom B1 — local winding-sensitive Hamiltonian

Let \(\nu\in\mathbb Z^7\) denote seven relative winding coordinates after removing a common phase. Define a bounded address score

\[
s(\nu)=\tanh\!\left(\frac{c\cdot\nu}{N_*}\right),
\]

where \(c\in\mathbb R^7\) and \(N_*>0\) must be fixed before data collection. Postulate

\[
H_{\mathrm{DR}}=H_{\mathrm{std}}+\frac{\hbar g}{2}s(\nu)\sigma_z.
\]

This is **not derived from the earlier mathematics**. It is the smallest bounded example that turns the winding hypothesis into a test.

For a Ramsey qubit with contrast \(C\), the model predicts

\[
P_+(\tau\mid\nu)=
\frac12\left[
1+C\cos\big((\omega_0+g\,s(\nu))\tau+\varphi\big)
\right].
\]

Two preparations \(A,B\) that are operationally identical under the standard state description but have different proposed windings predict

\[
\Delta\omega
=g\,[s(\nu_A)-s(\nu_B)].
\]

The null is

\[
H_0:g=0.
\]

The alternative is

\[
H_1:g\ne0
\]

with the sign reversals and scaling across \(\nu\) fixed in advance.

### Why this is a legitimate first test

- It supplies a number standard quantum mechanics does not contain if \(\nu\) is genuinely absent from the standard complete state.
- Conditional on a fixed local \(\nu\), the evolution remains linear and uses the ordinary Born rule.
- The interaction is local in this provisional version, reducing immediate no-signaling risk.
- It can be falsified or bounded.

### Why it does not yet prove the grand theory

- The form of \(s\), vector \(c\), scale \(N_*\), and operator \(\sigma_z\) are bridge assumptions, not deductions.
- An observed history effect could arise from calibration drift, hysteresis, non-Markovian noise, leakage, heating, or an incomplete standard state description.
- A null result constrains B1 and its parameter range; it does not disprove every metaphysical version of \(G\).
- A positive result would first establish unexplained memory, not automatically eight dimensions or a soul.

---

## XI. Preregistered experiment blueprint

### Experimental question

Can two preparations that end in the same certified standard physical state but differ in controlled closed-loop winding history produce different future Ramsey frequencies?

### Protocol

1. Choose a high-coherence qubit platform.
2. Randomly interleave preparation families \(A\), \(B\), and sham controls.
3. Make \(A\) and \(B\) traverse different closed control loops that the theory maps to \(\nu_A\) and \(\nu_B\).
4. Return both to the same target density matrix \(\rho_0\).
5. Certify operational equivalence using independent state tomography, leakage checks, pulse-energy matching, temperature monitoring, and history-sensitive process characterization.
6. Apply the identical Ramsey evolution and measurement sequence.
7. Blind the protocol labels until the analysis code and exclusion rules are frozen.
8. Estimate \(\Delta\omega\) jointly across several positive, negative, and zero values of the proposed winding difference.

### Primary endpoint

Use the preregistered regression

\[
\widehat{\omega}_j
=\omega_0+g\,s(\nu_j)+\beta^{\mathsf T}X_j+\epsilon_j,
\]

where \(X_j\) contains only prespecified nuisance monitors. The primary test is a confidence interval or Bayes factor for \(g\), with the evidentiary threshold chosen before unblinding.

### Required controls

- **Sham-loop control:** same duration and energy, zero claimed winding change.
- **Sign reversal:** \(\nu\to-\nu\) must reverse the predicted shift.
- **Magnitude series:** several \(|\nu|\) values must follow the frozen \(s(\nu)\).
- **Time reversal:** reverse the control loop where the model predicts reversal.
- **Platform replication:** repeat on a physically different qubit architecture.
- **No-signaling audit:** if remote systems are introduced, verify local marginals are independent of remote settings.
- **Standard-memory audit:** fit conventional non-Markovian and drift models before attributing any residual to \(\nu\).

### Falsification rule

B1 fails in the tested range if sufficiently precise data exclude the preregistered \(g\) range or if the effect fails its sign, magnitude, reversal, and cross-platform predictions. Allowing \(c\), \(N_*\), the functional form, or the sign to change after seeing results would convert the test into curve fitting.

### What must be derived next

The theory is not experiment-ready until it derives, rather than guesses:

1. why the state has eight raw phases;
2. which common phase is gauge and why;
3. how laboratory operations change \(\nu\);
4. the bridge operator \(A\);
5. the coupling function and scale;
6. the Born-rule or alternative outcome law;
7. Lorentz transformation properties;
8. a no-signaling proof;
9. a classical limit recovering Maxwellian, Newtonian/relativistic, and thermodynamic behavior where tested.

---

## XII. Reconstruction of the strongest neighboring experiments

The correct question is not whether an experiment sounds compatible with the framework. Almost any sufficiently elastic story can be made compatible after the fact. The question is whether the framework predicts a result that the complete standard model of the apparatus does not.

| Experiment class | Reconstructed standard prediction | What it does to the framework |
|---|---|---|
| Aharonov–Bohm loops | A charged wave acquires the gauge-invariant loop phase \(\Delta\phi=(q/\hbar)\oint \mathbf A\cdot d\mathbf l=q\Phi_B/\hbar\), even where the local magnetic field on the path vanishes | Shows that loop topology and winding can be physical, but already within ordinary gauge quantum mechanics |
| Berry/geometric phase | An adiabatically transported eigenstate acquires a path-dependent geometric phase in addition to its dynamical phase | Supports the usefulness of closed-loop geometry; does not establish an absolute winding memory |
| Delayed-choice quantum eraser | Entanglement creates a joint distribution whose coincidence-conditioned subsets can display complementary fringes; their unconditional sum has no fringe | No later human awareness is needed and no usable signal travels into the past |
| Non-Markovian process memory | An environment can retain past control information, so equal reduced density matrices need not have equal futures | Directly threatens B1: “same qubit state, different future” is not by itself new physics |
| Leggett–Garg tests | Temporal correlations can violate macrorealist/noninvasive bounds while remaining quantitatively quantum | Rules out a naive, noncontextual classical trajectory hidden behind the cycles |
| Massive matter-wave interference | Standard superposition persists for increasingly massive objects unless a modification supplies a quantitative collapse scale | Pressures any soul-, mass-, or complexity-triggered collapse claim |
| Multipath interference | With linear path amplitudes and the Born rule, all irreducible single-particle interference above order two vanishes | Supplies a direct structural null for an eight-way alignment coupling |

### 1. Closed paths are real physics, but not yet the missing variable

The magnetic Aharonov–Bohm phase was directly observed with the magnetic field shielded from the electron paths by Tonomura and collaborators [in 1986](https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.56.792). A multiply-orbiting trapped-ion interferometer later showed phase reversal when the direction of rotation was reversed [Saito and Mukaiyama (2024)](https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.133.143402). Controlled Berry phase has also been measured in a superconducting qubit [Leek et al. (2007)](https://arxiv.org/abs/0711.0218).

These results validate a restrained statement:

> Closed histories can carry measurable geometric or topological phase even when endpoints coincide.

They do **not** validate the stronger B1 statement that an absolute integer winding remains measurable after the complete standard physical state, gauge holonomy, environment, and control apparatus have all been matched. Aharonov–Bohm and Berry phases belong in \(Z_{\mathrm{std}}\), not in the unexplained residual.

### 2. Delayed choice does not make consciousness the causal variable

In the delayed-choice quantum eraser of Kim and collaborators, entangled signal–idler detections were sorted into coincidence-conditioned subsets [Kim et al. (2000)](https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.84.1). Schematically, two complementary conditioned patterns can be written

\[
p_{\pm}(x)=a(x)\left[1\pm V(x)\cos\phi(x)\right].
\]

If the idler result is ignored,

\[
p_{\mathrm{uncond}}(x)
=\tfrac12p_+(x)+\tfrac12p_-(x)
=a(x).
\]

The fringe and antifringe cancel. The later classification changes which correlations are displayed; it does not change the already recorded local marginal, and it does not require a person to become aware of the result. Any \(G\)-dependent observer theory must predict a different joint distribution or timing dependence, not merely redescribe this conditioning.

### 3. The process-tensor objection to “same state, different future”

The process-tensor framework gives an operational Markov criterion using interventions and causal breaks [Pollock et al. (2018)](https://arxiv.org/abs/1801.09811). An experiment on superconducting quantum processors reconstructed non-Markovian processes and used the reconstruction for improved prediction and control [White et al. (2020)](https://www.nature.com/articles/s41467-020-20113-3).

Let \(h_A,h_B\) be two controlled histories, \(\mathcal B_{\rho_0}\) a causal-break operation that discards the system and reprepares the same \(\rho_0\), and \(u\) a common future intervention. Define

\[
\delta_{\mathrm{hist}}(u)
=d_{\mathrm{TV}}\!\left[
P(Y\mid h_A,\mathcal B_{\rho_0},u),
P(Y\mid h_B,\mathcal B_{\rho_0},u)
\right].
\]

A nonzero \(\delta_{\mathrm{hist}}\) is evidence that history exists somewhere in the operational process. It is **not** automatically evidence for a new dimension: the environment, controller, leakage levels, temperature, or correlated noise can carry it.

The evidence ladder for B1 must therefore be:

1. equal state tomography;
2. equal leakage, energy, timing, and thermal monitors;
3. a system causal break;
4. out-of-sample prediction by a trained process-tensor or finite-memory model;
5. physical environment reset where available;
6. topological invariance across loop shape and speed;
7. orientation reversal under \(\nu\mapsto-\nu\);
8. replication on a different physical platform.

Even this ladder does not logically prove “dimension.” A sufficiently enlarged ordinary environment model can reproduce any finite history-dependent dataset. What could support B1 is the *predeclared invariant law* \(g\,s(\nu)\), especially across apparatuses—not the generic existence of memory.

### 4. Temporal realism and macroscopic collapse

Ideal-negative-measurement neutron interferometry reported a Leggett–Garg value \(K=1.120\pm0.007\) (statistical) \(\pm0.019\) (systematic), above the macrorealist bound, while detailed quantum dynamical modeling reproduced the result [Wagner et al. (2024)](https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.132.260201). Thus a hidden cyclic history cannot simply assign noninvasive, context-independent temporal values.

Matter-wave interference has now been demonstrated for sodium nanoparticles containing more than 7,000 atoms and exceeding \(170{,}000\) Da, with their center-of-mass superposition extending more than an order of magnitude beyond the particle diameter [Pedalino et al. (2026)](https://www.nature.com/articles/s41586-025-09917-9). The result found no mass- or size-only breakdown of superposition at that scale. It does not settle the measurement problem, but it makes every objective-collapse or consciousness-coupling proposal owe a quantitative mass, time, complexity, or record threshold.

### 5. Honest anomaly inventory

No reproducible result in the audited set presently requires B1, an eight-dimensional state, or a soul variable:

- Aharonov–Bohm and Berry phases are explained by standard gauge and geometric phase.
- Delayed-choice erasure is explained by joint quantum probabilities and postselection.
- History dependence is expected in open, non-Markovian processes.
- Leggett–Garg violations are predicted by quantum theory.
- Massive interference has so far strengthened, not broken, the standard superposition principle.
- Reported higher-order interference can be generated by standard nonlinear evolution or context-dependent path amplitudes.

The measurement problem remains conceptually open, but an interpretive problem is not the same thing as an unexplained numerical residual. At present the framework has candidate future tests, not a recovered anomaly that it uniquely explains.

---

## XIII. Bridge axiom B2 — an eight-path alignment interaction

### 1. The exact multipath hierarchy

Let \(N=\{1,\ldots,8\}\) label eight controllable paths. For every subset \(S\subseteq N\), let \(p_S(r)\) be the probability of output \(r\) when precisely the paths in \(S\) are enabled. Define the eighth-order Möbius, or Sorkin, term

\[
I_8(r)
=\sum_{S\subseteq N}(-1)^{8-|S|}p_S(r).
\]

More generally, for \(A\subseteq N\),

\[
I_A(r)
=\sum_{S\subseteq A}(-1)^{|A|-|S|}p_S(r).
\]

Under context-independent linear amplitudes and the Born rule,

\[
p_S(r)
=\left|\sum_{i\in S}\psi_i(r)\right|^2.
\]

#### Higher-order-null theorem

For every \(A\) with \(|A|\ge3\),

\[
I_A(r)=0.
\]

#### Proof

Expand

\[
p_S(r)
=\sum_{i\in S}|\psi_i(r)|^2
+\sum_{\substack{i,j\in S\\i\ne j}}
\psi_i(r)\psi_j(r)^*.
\]

As a function of the binary path indicators \(x_i=\mathbf1_{i\in S}\), this is a multilinear polynomial of degree at most two. The Möbius coefficient \(I_A\) extracts the coefficient of \(\prod_{i\in A}x_i\). No such monomial exists when \(|A|\ge3\), so the coefficient is zero. ∎

This is the hierarchy introduced by Sorkin’s generalized measure program [Sorkin (1994)](https://arxiv.org/abs/gr-qc/9401003). Third-order photon tests have bounded deviations from the Born hierarchy [Sinha et al. (2010)](https://arxiv.org/abs/1007.4193), a single-spin implementation reached the \(10^{-3}\) scale [Jin et al. (2017)](https://arxiv.org/abs/1612.08563), and a four-path photonic chip reported \(\kappa=-0.0031\pm0.0047\) [Chen et al. (2021)](https://www.nature.com/articles/s41467-021-22887-6).

### 2. A minimal eight-way bridge

For wrapped path phases \(\theta=(\theta_1,\ldots,\theta_8)\), define the toroidal alignment defect

\[
\Delta_{\mathbb T,8}(\theta)
=\min_{\alpha\in[0,2\pi)}
\left[
\sum_{i=1}^8
\operatorname{wrap}(\theta_i-\alpha)^2
\right]^{1/2}.
\]

This vanishes when all eight phases agree up to common phase, precisely the seven-relative-phase version of the alignment line. Define a fixed gate

\[
G_\sigma(\theta)
=\exp\!\left[-\frac{\Delta_{\mathbb T,8}(\theta)^2}{2\sigma^2}\right],
\]

with \(\sigma\) fixed before data collection.

Bridge axiom B2 postulates, for output channels \(r=1,\ldots,8\),

\[
p_S^{\mathrm{DR}}(r\mid\theta)
=p_S^{\mathrm{QM}}(r\mid\theta)
+\varepsilon\,
\mathbf1_{S=N}\,
G_\sigma(\theta)\,a_r,
\]

where

\[
\sum_{r=1}^8 a_r=0
\]

preserves normalization, and \(\varepsilon\) is small enough to preserve positivity. The vector \(a\), gate width \(\sigma\), and sign convention must be frozen from symmetry or calibration-independent apparatus geometry before unblinding.

Then

\[
I_8^{\mathrm{DR}}(r\mid\theta)
=\varepsilon G_\sigma(\theta)a_r,
\]

while the added term contributes nothing to any experiment with fewer than all eight paths.

B2 is not derived from the projection theorem. It is a deliberately exposed bridge assumption. Its value is that it produces a distinctive conjunction:

- \(I_3,\ldots,I_7\) remain null after corrections;
- \(I_8\) follows the frozen alignment envelope \(G_\sigma\);
- disabling any single path kills the term;
- phase scrambling suppresses it;
- the normalized effect is independent of optical intensity in the single-particle regime.

### 3. The decisive loopholes

A nonzero naive \(I_8\) is not sufficient. Opening or closing paths can alter boundary conditions and therefore change other path amplitudes; nonclassical looped trajectories can make a naive Sorkin parameter nonzero within ordinary quantum electrodynamics [Sinha et al. (2017)](https://www.nature.com/articles/ncomms13987). Standard nonlinear evolution also produces higher-order interference [Namdar et al. (2023)](https://journals.aps.org/pra/abstract/10.1103/PhysRevA.107.032211).

The polynomial degree exposes the nonlinear loophole. If

\[
p_S\propto\left|\sum_{i\in S}\psi_i\right|^{2q},
\]

then the set function can contain terms through order \(2q\). A fourth-power response can fake \(I_3\) and \(I_4\), while an eighth-power response can fake \(I_8\). Therefore the experiment must establish source and detector linearity, repeat over several flux decades, and fit measured nonlinear transfer functions before forming the residual hierarchy.

### 4. Proposed eight-path experiment

A practical platform is a phase-stable time-bin or integrated-photonic interferometer. Eight-dimensional temporal circuits have already implemented hundreds of programmable unitaries with reported fidelities above 97% [Bouchard et al. (2024)](https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.133.090601), while programmable time-bin work has observed interference of up to eight photons in sixteen modes with one detector [Carosini et al. (2024)](https://www.science.org/doi/10.1126/sciadv.adj0993). Those are enabling technologies, not tests of B2.

The protocol should be:

1. Prepare a heralded single photon coherently over eight time bins or paths.
2. Randomize all \(2^8=256\) enabled-path subsets shot block by shot.
3. Scan a preregistered grid through \(\Delta_{\mathbb T,8}=0\) and phase-scrambled controls.
4. Estimate the complete transfer matrix and detector response independently for each switching context.
5. Compute \(I_A(r)\) for every order \(3\le |A|\le8\), not only the desired \(I_8\).
6. Fit Maxwell/quantum, switching-crosstalk, dark-count, drift, multiphoton, and detector-nonlinearity models before examining the B2 residual.
7. Require the full order–phase–intensity fingerprint, not a single significant bin.
8. Replicate with a different switching mechanism and at an independent laboratory.

The cleanest evidentiary comparison is:

\[
\begin{array}{ll}
M_0:&\text{linear quantum model plus calibrated apparatus},\\
M_1:&\text{standard nonlinear/context-dependent apparatus model},\\
M_2:&M_1+\varepsilon G_\sigma(\theta)a_r\mathbf1_{S=N}.
\end{array}
\]

A discovery claim requires out-of-sample preference for \(M_2\), the predicted hierarchy fingerprint, independent replication, and a demonstrated failure of plausible \(M_1\) variants. Even then, it would establish a new eight-way interaction—not by itself a soul or eight spacetime dimensions.

---

## XIV. Statistical power and rejection thresholds

The following numbers use a two-sided five-standard-deviation discovery threshold and 90% power:

\[
z_*=5+\Phi^{-1}(0.90)=6.28155.
\]

They are planning bounds, not substitutes for a full likelihood with nuisance parameters, drift, multiple testing, and sequential stopping.

### 1. Two matched histories

For two independent Bernoulli outcomes with worst-case probability \(p=1/2\) and an absolute probability difference \(d\), the normal approximation gives

\[
n_{\mathrm{per\ history}}
\approx
\frac{2z_*^2p(1-p)}{d^2}
=\frac{19.729}{d^2}.
\]

| Absolute history effect \(d\) | Shots per history |
|---:|---:|
| 0.020 | 49,323 |
| 0.010 | 197,290 |
| 0.005 | 789,158 |
| 0.001 | 19,728,946 |

### 2. Ramsey winding shift

At quadrature, for Ramsey contrast \(C\) and differential accumulated phase \(\delta=\Delta\omega\,\tau\),

\[
n_{\mathrm{per\ history}}
\approx
\frac{2z_*^2}{C^2\delta^2}.
\]

For \(C=0.9\):

| Differential phase \(|\delta|\) | Shots per history |
|---:|---:|
| 0.020 rad | 243,568 |
| 0.010 rad | 974,269 |
| 0.005 rad | 3,897,076 |
| 0.001 rad | 97,426,890 |

These counts ignore phase-scan inefficiency, frequency fitting, nuisance estimation, and familywise error across several winding values, so a real design requires more.

### 3. Direct eighth-order inclusion–exclusion

If every \(p_S(r)\) is estimated from \(N\) independent Bernoulli shots and no covariance reduction is available, then

\[
\operatorname{Var}(\widehat I_8)
=\sum_{S\subseteq N}\operatorname{Var}(\widehat p_S)
\le\frac{256}{4N}
=\frac{64}{N}.
\]

Thus the worst-case standard error is \(8/\sqrt N\), and detecting an absolute eighth-order term \(\epsilon\) requires approximately

\[
N_{\mathrm{per\ configuration}}
\ge
\left(\frac{8z_*}{\epsilon}\right)^2
=\frac{2525.3}{\epsilon^2}.
\]

| Absolute \(I_8\) signal \(\epsilon\) | Shots/configuration | Total over 256 configurations |
|---:|---:|---:|
| 0.050 | 1,010,122 | 258,591,232 |
| 0.020 | 6,313,263 | 1,616,195,328 |
| 0.010 | 25,253,050 | 6,464,780,800 |
| 0.005 | 101,012,199 | 25,859,122,944 |
| 0.001 | 2,525,304,965 | 646,478,071,040 |

This conservative calculation reveals a serious weakness: direct eighth-order subtraction amplifies noise exponentially. High-rate photonics, common-mode covariance, balanced Walsh coding, and a preregistered parametric phase envelope can improve efficiency, but every improvement adds assumptions. A staged program should first validate the entire analysis at orders three and four, then proceed to eight only when systematic residuals are already below the targeted \(\epsilon\).

### 4. Computational theorem check

As an implementation audit, 2,000 random complex-amplitude trials were run separately for each path count \(n=3,\ldots,8\). For \(p_S=|\sum_{i\in S}\psi_i|^2\), the largest computed \(|I_n|\) across all 12,000 trials was below \(1.1\times10^{-14}\), consistent with floating-point roundoff and the theorem. Replacing the response by \(|\sum\psi_i|^4\) produced nonzero third- and fourth-order terms; replacing it by \(|\sum\psi_i|^8\) produced nonzero eighth-order terms. This verifies the implementation and the nonlinear loophole. It is not empirical evidence about nature.

### 5. The missing number is now explicit

Neither B1 nor B2 predicts a numerical magnitude:

\[
g=\;?\qquad\text{and}\qquad \varepsilon=\;?
\]

Without at least a theoretically motivated range, no experiment can be declared adequately powered in advance; a null can always be evaded by making the coupling smaller. Deriving \(g\) or \(\varepsilon\) from independently fixed constants is now the highest-value mathematical task.

---

## XV. Dimensional lift into quantum computation

### 1. The bird’s-eye verdict

There is an exact quantum-computational object hiding inside the eight-coordinate construction:

\[
\boxed{
\text{eight basis phases}
\longleftrightarrow
\text{one global phase}
+\text{seven relative interaction coordinates}
}
\]

The correspondence is rigorous, invertible for a chosen unwrapped phase lift, and experimentally realizable on three qubits. It yields:

- an exact phase compiler;
- an exact alignment detector;
- closed formulas for entanglement;
- an exact eight-vertex quantum walk;
- a universal gate route when noncommuting Hadamard operations are added.

It does **not** prove eight spacetime dimensions or a novel quantum law. Walsh synthesis of diagonal unitaries and multi-qubit hypergraph phase states are established quantum-information structures [Welch et al. (2014)](https://arxiv.org/abs/1306.3991), [Rossi et al. (2013)](https://arxiv.org/abs/1211.5554). The possible contribution is a disciplined synthesis: use dimensional lifting, projection closure, winding canonicalization, and sparse interaction discovery as one compiler and diagnostic framework.

### 2. Four objects previously called “zero”

Quantum theory forces four ideas to be separated:

| Object | Formula | Meaning |
|---|---|---|
| Vector zero | \(0\in\mathbb C^8\) | Not a legal quantum state because its norm is zero |
| Computational zero | \(|000\rangle\) | A normalized basis state |
| Uniform potential state | \(|+\!+\!+\rangle=8^{-1/2}\sum_x|x\rangle\) | Equal measurement weight over eight basis alternatives |
| Phase alignment | \(D(\theta)=e^{i\alpha}I\) | All eight phases agree; the phase layer changes nothing observable |

Thus “zero contains all potential” has a clean operational replacement:

\[
\text{prepare }|000\rangle
\xrightarrow{H^{\otimes3}}
|+\!+\!+\rangle.
\]

The zero vector does not contain possibilities. The normalized uniform superposition carries coherent amplitude over the eight computational alternatives.

### 3. State-space dimension audit

A general pure three-qubit state is

\[
|\psi\rangle=\sum_{x=0}^{7}c_x|x\rangle,
\qquad c_x\in\mathbb C,
\qquad \sum_x|c_x|^2=1.
\]

Eight complex amplitudes contain sixteen real numbers. Normalization removes one and global phase removes one, leaving

\[
\dim_{\mathbb R}\mathbb{CP}^{7}=14.
\]

The eight-phase model restricts to equal magnitudes:

\[
|\psi(\theta)\rangle
=\frac1{\sqrt8}\sum_{x=0}^{7}e^{i\theta_x}|x\rangle.
\]

It has eight phase coordinates, but the common shift

\[
\theta_x\mapsto\theta_x+\alpha
\]

multiplies the whole state by an unobservable global phase. Therefore its physical dimension is

\[
8-1=7.
\]

This exactly realizes the earlier quotient

\[
\mathbb T^8/S^1\cong\mathbb T^7.
\]

But it is only a seven-dimensional submanifold of the full fourteen-dimensional pure-state space. To represent arbitrary three-qubit pure states, one needs seven independent probability coordinates in the simplex plus seven relative phases:

\[
(p_0,\ldots,p_7)\in\Delta_7,
\qquad
(\theta_1-\theta_0,\ldots,\theta_7-\theta_0)\in\mathbb T^7,
\]

away from boundaries where some \(p_x=0\) and the corresponding phase is undefined.

### 4. Exact Walsh/Pauli dimensional lift

Let \(x=(x_1,x_2,x_3)\in\{0,1\}^3\). For each subset \(S\subseteq\{1,2,3\}\), define the Walsh character

\[
\chi_S(x)=(-1)^{\sum_{j\in S}x_j}.
\]

Every unwrapped real phase table \(\theta_x\) has the unique expansion

\[
\theta_x=\sum_{S\subseteq\{1,2,3\}}\alpha_S\chi_S(x),
\]

with inverse

\[
\boxed{
\alpha_S=\frac18\sum_{x\in\{0,1\}^3}\theta_x\chi_S(x)
}.
\]

Orthogonality,

\[
\frac18\sum_x\chi_S(x)\chi_T(x)=\delta_{ST},
\]

proves both uniqueness and exact inversion.

Let

\[
Z_S=\prod_{j\in S}Z_j,
\qquad Z_\varnothing=I.
\]

Because \(Z_S|x\rangle=\chi_S(x)|x\rangle\), the diagonal phase unitary

\[
D(\theta)=\sum_xe^{i\theta_x}|x\rangle\langle x|
\]

factorizes exactly:

\[
\boxed{
D(\theta)
=\exp\!\left(i\sum_S\alpha_SZ_S\right)
=e^{i\alpha_\varnothing}
\prod_{S\ne\varnothing}e^{i\alpha_SZ_S}
}.
\]

The eight lifted coordinates have a precise hierarchy:

| Count | Coordinate type | Operator | Role |
|---:|---|---|---|
| 1 | Global | \(I\) | Unobservable gauge phase |
| 3 | One-body | \(Z_1,Z_2,Z_3\) | Local phase rotations |
| 3 | Two-body | \(Z_1Z_2,Z_1Z_3,Z_2Z_3\) | Pair interactions, potentially entangling |
| 1 | Three-body | \(Z_1Z_2Z_3\) | Collective interaction, potentially genuinely tripartite |

This is the exact mathematical pattern behind “one plus seven.” The eighth raw coordinate is the common phase; the seven remaining coordinates are not seven extra spaces but seven observable diagonal interaction channels.

### 5. The winding correction

The physical unitary is unchanged under

\[
\theta_x\mapsto\theta_x+2\pi n_x,
\qquad n_x\in\mathbb Z.
\]

Consequently, applying the real Walsh transform directly to wrapped principal phases is not a unique compiler. Different integer lifts produce different coefficient vectors that represent the same unitary:

\[
\alpha_S(n)
=\frac18\sum_x(\theta_x+2\pi n_x)\chi_S(x).
\]

After the global component is removed, these shifts form the Walsh equivalence lattice

\[
\Lambda_W
=\left\{
\left(
\frac{2\pi}{8}\sum_xn_x\chi_S(x)
\right)_{S\ne\varnothing}
:n\in\mathbb Z^8
\right\}.
\]

Thus the physical coefficient space is \(\mathbb R^7/\Lambda_W\), not unconstrained \(\mathbb R^7\).

A compiler must choose a representative by a frozen rule, for example

\[
n^*
=\arg\min_{n\in\mathbb Z^8}
\mathcal C\!\left(\alpha(n)\right),
\]

where \(\mathcal C\) is a preregistered gate-cost function. This is a lattice/canonicalization problem. It is also the exact boundary between standard and new winding:

- if every lift produces the same operational unitary, winding is compiler redundancy;
- if nature produces a further \(n\)-dependent observable, that is the separate B1 physical hypothesis.

### 6. Projection closure becomes experimentally exact

Suppose the observer retains only computational-basis probabilities:

\[
\pi_Z(|\psi\rangle)
=\bigl(|\langle x|\psi\rangle|^2\bigr)_{x=0}^{7}.
\]

For every phase state above,

\[
\pi_Z(|\psi(\theta)\rangle)
=\left(\frac18,\ldots,\frac18\right).
\]

The projection loses all seven relative phases. This loss is maximally predictive, not philosophical.

For a nonzero bit mask \(m\), define

\[
|\psi_m\rangle
=\frac1{\sqrt8}\sum_x(-1)^{m\cdot x}|x\rangle.
\]

Every \(|\psi_m\rangle\) has the same current probability vector. Yet

\[
H^{\otimes3}|\psi_m\rangle=|m\rangle.
\]

Therefore two states in the same probability fiber can evolve under the same future operation into disjoint deterministic outcomes. Their total-variation distance is one:

\[
d_{\mathrm{TV}}(\delta_m,\delta_{m'})=1
\qquad(m\ne m').
\]

Hence, for this observation map and allowed Hadamard intervention,

\[
\boxed{\mathcal D(\pi_Z)=1}.
\]

This proves a central intuition of the framework inside standard quantum mechanics: the observed coordinate can be maximally incomplete even when its present probability description looks identical.

### 7. Alignment detector and phase decoder

Prepare \(|+\!+\!+\rangle\), apply \(D(\theta)\), apply \(H^{\otimes3}\), and measure. The exact output law is

\[
\boxed{
P(y\mid\theta)
=\left|
\frac18\sum_{x\in\{0,1\}^3}
(-1)^{x\cdot y}e^{i\theta_x}
\right|^2
}.
\]

The probability of output \(000\) is

\[
P(000\mid\theta)
=\left|\frac18\sum_xe^{i\theta_x}\right|^2
\equiv F_{\mathrm{align}}.
\]

This quantity is invariant under common phase and satisfies

\[
0\le F_{\mathrm{align}}\le1,
\]

with

\[
F_{\mathrm{align}}=1
\iff
\theta_0=\cdots=\theta_7\pmod{2\pi}.
\]

Thus the circuit is an exact physical measurement of the toroidal alignment condition.

For small gauge-fixed deviations \(\delta_x\) with \(\sum_x\delta_x=0\),

\[
F_{\mathrm{align}}
=1-\frac18\sum_x\delta_x^2+O(\|\delta\|^4).
\]

So the earlier geometric alignment defect becomes, locally,

\[
1-F_{\mathrm{align}}
\approx\frac{\Delta_{\mathbb T,8}^2}{8}.
\]

For a promised Walsh character,

\[
e^{i\theta_x}=(-1)^{m\cdot x},
\]

the output is deterministic:

\[
P(y\mid\theta)=\delta_{y,m}.
\]

This is a clean phase-to-bit decoder closely related to the phase-oracle structure of the Deutsch–Jozsa and Bernstein–Vazirani algorithms [Deutsch and Jozsa (1992)](https://royalsocietypublishing.org/doi/10.1098/rspa.1992.0167), [Bernstein and Vazirani (1997)](https://epubs.siam.org/doi/10.1137/S0097539796300921). It is powerful under a promise; it does not reveal an arbitrary eight-entry phase table from one sample.

### 8. Exact entanglement prediction from the phase geometry

For qubit \(j\), define the conditional phase overlap

\[
\eta_j
=\frac14
\sum_{r\in\{0,1\}^2}
\exp\!\left[
i\bigl(\theta_{0r}^{(j)}-\theta_{1r}^{(j)}\bigr)
\right].
\]

The one-qubit reduced state is

\[
\rho_j
=\frac12
\begin{pmatrix}
1&\eta_j\\
\eta_j^*&1
\end{pmatrix},
\]

and its purity is

\[
\boxed{
\operatorname{Tr}(\rho_j^2)
=\frac{1+|\eta_j|^2}{2}
}.
\]

For the global pure state:

- \(\operatorname{Tr}(\rho_j^2)=1\) iff qubit \(j\) is separable from the other two;
- \(\operatorname{Tr}(\rho_j^2)<1\) iff it is entangled across the \(j\mid\overline j\) cut;
- all three purities below one imply genuine tripartite entanglement.

This gives an exact entanglement predictor directly from the eight phases. The interaction-order table suggests where to look, but coefficient order alone is not sufficient because special angles can cancel entanglement. Purity performs the actual test.

Canonical examples at angle \(\pi/4\) are:

| Phase generator | Hadamard outputs | Single-qubit purities | Verdict |
|---|---|---|---|
| \(I\) | \(P(000)=1\) | \((1,1,1)\) | Aligned, separable |
| \(Z_1\) | \(P(000)=P(001)=1/2\) | \((1,1,1)\) | Local phase, separable |
| \(Z_1Z_2\) | \(P(000)=P(011)=1/2\) | \((1/2,1/2,1)\) | Pair entanglement |
| \(Z_1Z_2Z_3\) | \(P(000)=P(111)=1/2\) | \((1/2,1/2,1/2)\) | Genuine tripartite entanglement |

The phase table with only \(\theta_{111}=\pi\) is the three-qubit CCZ hypergraph state. The audited calculation gives

\[
F_{\mathrm{align}}=\frac9{16},
\qquad
\operatorname{Tr}(\rho_j^2)=\frac58
\quad\text{for every }j.
\]

Hypergraph states and their multipartite entanglement are established prior art, so this is an exact correspondence, not a new discovery by itself [Rossi et al. (2013)](https://arxiv.org/abs/1211.5554).

### 9. Exact quantum movement on the eight-vertex cube

The phase layer alone does not move computational-basis probability:

\[
[D(\theta),|x\rangle\langle x|]=0.
\]

Actual amplitude transport requires off-diagonal generators. Place the eight basis states at the vertices of the Boolean cube and choose

\[
H_{\mathrm{move}}
=\hbar\kappa(X_1+X_2+X_3).
\]

Starting from \(|000\rangle\),

\[
|\psi(t)\rangle
=\bigl(
\cos(\kappa t)|0\rangle
-i\sin(\kappa t)|1\rangle
\bigr)^{\otimes3}.
\]

For vertex \(x\) of Hamming weight \(w(x)\),

\[
\boxed{
P_t(x)
=\cos^{2(3-w(x))}(\kappa t)
\sin^{2w(x)}(\kappa t)
}.
\]

This predicts three diagnostic moments exactly:

\[
\begin{array}{c|c}
\kappa t&\text{distribution}\\ \hline
0&P(000)=1\\
\pi/4&P(x)=1/8\text{ for all eight vertices}\\
\pi/2&P(111)=1
\end{array}
\]

This walk is separable because its Hamiltonian is a sum of independent one-body terms. Add noncommuting pair or triple interactions to create entanglement. The distinction is essential:

> probability motion requires off-diagonal coupling; entanglement requires nonseparable interaction; phase accumulation alone is neither.

### 10. Does this give a quantum computer?

It gives a mathematically clean three-qubit phase processor. For universal quantum computation, a model needs:

1. normalized state preparation;
2. coherent operations forming a universal gate set;
3. tensor-product composition;
4. measurement probabilities;
5. an error model and correction procedure;
6. polynomial resource accounting for the claimed algorithm;
7. a physical realization of the operations.

The diagonal lift supplies arbitrary \(Z_S\) phase interactions on three qubits. Independently addressable Hadamards add a noncommuting basis change:

\[
X_j=H_jZ_jH_j.
\]

Arbitrary single-qubit operations plus CNOT are universal [Barenco et al. (1995)](https://arxiv.org/abs/quant-ph/9503016). More discretely, Hadamard, the \(\pi/8\) phase gate, and CNOT form a universal fault-tolerant basis [Boykin et al. (1999)](https://arxiv.org/abs/quant-ph/9906054). Therefore the dimensional phase compiler can sit inside a universal architecture.

But one seven-parameter diagonal layer is not an arbitrary three-qubit unitary:

\[
\dim SU(8)=8^2-1=63,
\]

whereas the physical diagonal layer has only seven parameters. Universality comes from composing it with noncommuting gates, not from the number eight alone.

### 11. Scalability and the real pattern-search opportunity

For \(n\) qubits, an arbitrary diagonal unitary has

\[
2^n-1
\]

physical phases. Its exact Walsh expansion can therefore be exponentially large. Merely storing an arbitrary phase table already costs exponential resources; a classical simulator using complex double amplitudes similarly uses

\[
16\cdot2^n\text{ bytes}.
\]

The compiler is efficient only when the phase function has structure. Define its Walsh support

\[
\operatorname{supp}_W(\theta)
=\{S:\alpha_S\ne0\}
\]

and sparsity

\[
s_W(\theta)=|\operatorname{supp}_W(\theta)|.
\]

If \(s_W(\theta)\) and the interaction weights \(|S|\) are polynomially bounded, the phase operation has a compact local description and can be compiled efficiently. If a generic table has exponentially many significant coefficients, the lift diagnoses rather than removes the exponential complexity.

This is the defensible pattern-search target:

\[
\boxed{
\text{find a low-order, sparse interaction language}
\quad\text{instead of memorizing }2^n\text{ phases}.
}
\]

It mirrors the earlier relation-lift discipline: recognize an exact algebra, synthesize a compact operator, verify equivalence, then compile. Walsh synthesis is known prior art; a new result would require a better canonical cost, a provable sparsity theorem for a useful problem class, or a demonstrable reduction in gates/noise over existing compilers.

### 12. Noise law for the alignment decoder

Let the eight branches receive independent zero-mean Gaussian phase errors

\[
\epsilon_x\sim\mathcal N(0,\sigma^2).
\]

For an aligned state or a promised character decoded to its target output,

\[
\mathbb E[P_{\mathrm{correct}}]
=\frac18+\frac78e^{-\sigma^2}.
\]

The derivation separates eight diagonal terms from \(8\cdot7\) off-diagonal terms and uses

\[
\mathbb E[e^{i(\epsilon_x-\epsilon_y)}]=e^{-\sigma^2}.
\]

This produces an immediate engineering prediction: phase noise does not merely “blur dimensions”; it reduces the coherent cross terms exponentially in its variance and leaves the incoherent floor \(1/8\).

Fault-tolerant scalability requires more than this toy noise law. The quantum threshold theorem proves that arbitrarily long computation is possible with controlled overhead under stated locality/noise assumptions when physical error lies below a positive threshold [Aharonov and Ben-Or (1999)](https://arxiv.org/abs/quant-ph/9906129). The present framework must specify its noise channel before quoting any hardware threshold.

### 13. Frozen dimensional-quantum protocol Q1

The following is now the canonical, falsifiable computation layer:

1. **State domain:** three-qubit equal-magnitude phase states.
2. **Gauge:** common phase is removed.
3. **Lift:** compute the eight Walsh coefficients \(\alpha_S\).
4. **Hierarchy:** classify global, local, pair, and triple interactions by \(|S|\).
5. **Compile:** implement \(\prod_{S\ne\varnothing}e^{i\alpha_SZ_S}\).
6. **Move:** add off-diagonal \(X/Y\) terms or Hadamard layers when probability transport is required.
7. **Predict:** calculate the full Born distribution, not an individual unmeasured outcome.
8. **Entanglement:** calculate all single-qubit purities; do not infer entanglement from visual complexity.
9. **Canonicalize winding:** freeze the integer phase lift by a declared gate-cost function.
10. **Verify:** compare statevector, compiled circuit, and sampled frequencies with held-out settings.
11. **Scale audit:** count state preparation, oracle construction, gates, depth, shots, and classical preprocessing.
12. **Novelty gate:** do not call standard Walsh synthesis, hypergraph states, or Hadamard decoding a new law.

### 14. What can and cannot be predicted

The framework can now predict, exactly within Q1:

- amplitude transport under a specified Hamiltonian;
- the entire measurement distribution after any compiled three-qubit phase circuit;
- phase alignment fidelity;
- entanglement across each one-versus-two qubit cut;
- response to the stated Gaussian phase-noise model;
- gate structure implied by the seven lifted interaction coordinates.

It cannot presently predict:

- the unique result of a genuinely quantum measurement before measurement;
- an unknown Hamiltonian without calibration or data;
- inaccessible “absolute winding” effects without B1;
- non-Born outcomes without a new probability law;
- scalable advantage for arbitrary exponentially specified phase tables;
- a hardware material, control pulse, or error threshold from the number eight.

If an additional hidden variable \(\lambda\) is proposed to predict individual outcomes, it must satisfy

\[
P(y\mid C)
=\int P(y\mid C,\lambda)\,d\mu(\lambda)
=P_{\mathrm{Born}}(y\mid C)
\]

for established experiments, while producing preregistered accessible predictive information on held-out trials. Bell and contextuality results still prohibit a simple local, noncontextual predetermined-outcome account.

### 15. Court-grade conclusion of the lift

The lift succeeds mathematically:

\[
\boxed{
\mathbb T^8/S^1\cong\mathbb T^7
\xrightarrow[\text{exact}]{\text{Walsh coordinates}}
\mathbb R^7/\Lambda_W\text{ diagonal Pauli interactions}
\xrightarrow{H\text{ and entanglers}}
\text{universal quantum circuit language}.
}
\]

The strongest finding is not “eight dimensions build a quantum computer.” It is:

> The eight-coordinate phase geometry is exactly a three-qubit diagonal phase sector. Its seven relative coordinates are a complete interaction basis for that sector; Hadamard interference turns them into measurable output, and reduced purities turn them into exact entanglement predictions.

This is a clean mathematical architecture and a useful compiler target. Its physics is standard until B1 or B2 produces a replicated residual beyond the complete standard model.

---

## XVI. The Born-outcome prediction frontier

### 1. Verdict before construction

There is now a rigorous method for **searching** for an accessible predictor of individual measurement outcomes. There is not yet evidence that such a predictor exists in nature. This distinction is the frontier wall.

Four achievements must not be conflated:

| Level | Achievement | Scientific meaning |
|---|---|---|
| BPO-0 | Calculate \(P_{\mathrm{Born}}(y\mid C)\) | Standard quantum prediction of a distribution |
| BPO-1 | Predict apparatus drift or memory omitted from \(C\) | Better engineering model, not automatically new physics |
| BPO-2 | Use a pre-measurement variable to predict held-out individual outcomes after the complete standard context is included | A genuine breach of the stated Born conditional model |
| BPO-3 | Replicate BPO-2 across devices while satisfying Bell, contextuality, no-signaling, relativistic, and independence constraints | Candidate new physical completion |

The present work reaches a theorem and protocol for distinguishing these levels. It does **not** claim BPO-2 or BPO-3.

### 2. The Born Prediction Wall theorem

At trial \(t\), let \(\mathcal F_{t-1}\) contain everything legitimately available from earlier trials. Let \(C_t\) contain the declared standard context: prepared quantum state, measurement operators, setting, calibrated control variables, and every standard history variable included by the experimental model. Quantum theory gives

\[
p_t(y)=P_{\mathrm{Born}}(Y_t=y\mid C_t,\mathcal F_{t-1})
=\operatorname{Tr}(E_{t,y}\rho_t).
\]

A candidate computational method outputs a normalized distribution \(q_t(y)\) before the outcome is known. It may be retrained using \(\mathcal F_{t-1}\), but it may not use \(Y_t\) or information causally downstream from \(Y_t\).

**Theorem BPW.** If the Born conditional law is correct, no such predictor has positive expected log-score advantage:

\[
\begin{aligned}
\mathbb E_{p_t}\!\left[
\log\frac{q_t(Y_t)}{p_t(Y_t)}
\middle|C_t,\mathcal F_{t-1}
\right]
&=\sum_y p_t(y)\log\frac{q_t(y)}{p_t(y)}\\
&=-D_{\mathrm{KL}}(p_t\Vert q_t)\le 0.
\end{aligned}
\]

Equality holds only when \(q_t=p_t\) on the support of \(p_t\). This is an information wall, not a computing-power wall: faster optimization cannot extract predictive information that is absent from the predictor's inputs.

**Proof.** The displayed expectation is the negative definition of Kullback–Leibler divergence, which is nonnegative by the log-sum inequality. No independence or identical-distribution assumption across trials is required once the conditional law is correctly specified. \(\square\)

The corresponding likelihood-ratio process is

\[
E_T=\prod_{t=1}^{T}\frac{q_t(Y_t)}{p_t(Y_t)},\qquad E_0=1.
\]

Under the null,

\[
\mathbb E\!\left[\frac{E_t}{E_{t-1}}\middle|C_t,\mathcal F_{t-1}\right]
=\sum_y p_t(y)\frac{q_t(y)}{p_t(y)}
=1.
\]

Therefore \(E_T\) is a nonnegative martingale under the simple Born null. Ville's inequality gives the anytime-valid guarantee

\[
P_{\mathrm{Born}}\!\left(\sup_T E_T\ge \frac1\alpha\right)\le\alpha.
\]

This allows continuous monitoring without manufacturing significance by stopping at a lucky moment. Modern e-value methods formalize this safe optional continuation [Grünwald, de Heide, and Koolen (2024)](https://academic.oup.com/jrsssb/article/86/5/1091/7617024); time-uniform martingale methods give the broader sequential-inference foundation [Howard et al. (2020)](https://arxiv.org/abs/1808.03204).

### 3. The exact breach condition

Let \(\Lambda_t\) be an accessible variable recorded before the measurement outcome. The wall is informationally breached exactly when

\[
\boxed{I(Y_t;\Lambda_t\mid C_t,\mathcal F_{t-1})>0}
\]

on genuinely held-out trials. If the optimal lifted predictor is the true conditional law

\[
q_t^*(y)=P(Y_t=y\mid \Lambda_t,C_t,\mathcal F_{t-1}),
\]

then its expected log-score gain over the complete context-only distribution is precisely

\[
\mathbb E\!\left[\log\frac{q_t^*(Y_t)}{p_t(Y_t)}\right]
=I(Y_t;\Lambda_t\mid C_t,\mathcal F_{t-1}).
\]

This supplies the missing variable's operational definition: it is not “a hidden dimension” by declaration; it is a pre-outcome, accessible state component with reproducible conditional mutual information after known physics and apparatus state have been controlled.

Positive mutual information is still much weaker than deterministic prediction. Perfect prediction requires

\[
H(Y_t\mid \Lambda_t,C_t,\mathcal F_{t-1})=0.
\]

The first scientifically meaningful target is therefore any replicated \(I>0\), not an immediate claim of certainty.

### 4. The Dimensional Outcome Predictor

The Q1 phase compiler remains the standard-physics baseline. Its seven relative Walsh/Pauli coordinates \(\alpha_S\), calibrated Hamiltonian, noise model, and measurement setting belong inside \(C_t\). Reusing one of those known phase variables as if it were new would only rediscover the Born distribution.

For a candidate dimensional lift \(L_k\) acting on the pre-measurement history \(h_t\), define

\[
\Lambda_{t,k}=L_k(h_t).
\]

Candidate families can include:

- the canonical integer winding \(\nu_t\) left after wrapped relative phase is specified;
- predictive-state coordinates inferred from control and sensor histories;
- process-memory coordinates surviving causal breaks and device resets;
- a declared B1 or B2 interaction statistic;
- negative-control coordinates that should have no predictive power.

For outcomes \(y\), turn a candidate score \(f_{k,y}\) into a normalized prediction by an exponential tilt of the Born baseline:

\[
\boxed{
q_{t,k}(y\mid \Lambda_{t,k},C_t)
=\frac{p_t(y\mid C_t)e^{f_{k,y}(\Lambda_{t,k},C_t)}}
{\sum_z p_t(z\mid C_t)e^{f_{k,z}(\Lambda_{t,k},C_t)}}.
}
\]

For a binary outcome this becomes

\[
q_{t,k}(1)=\sigma\!\left(
\operatorname{logit}p_t(1)+f_k(\Lambda_{t,k},C_t)
\right).
\]

Setting \(f=0\) recovers Born exactly. The lift is trained only on past trials, frozen for the current prediction, and scored on the next unrevealed outcome. This is the direct computational use of projection closure: generate candidate history lifts, ask which one closes the future outcome law, and reject every lift that fails prequential prediction.

### 5. Safe search across many possible lifts

Searching thousands of dimensional features and reporting only the lucky winner would invalidate the evidence. Predeclare candidate predictors \(q_{t,1},\ldots,q_{t,K}\), give them nonnegative weights \(w_k\) summing to one, and compute

\[
E_{T,k}=\prod_{t=1}^{T}\frac{q_{t,k}(Y_t)}{p_t(Y_t)},
\qquad
E_T^{\mathrm{mix}}=\sum_{k=1}^{K}w_kE_{T,k}.
\]

The mixture is itself a valid e-value because expectation is linear. It lets the computation search a fixed library of phase, winding, memory, and interaction lifts without a post-selection loophole. New candidate families require new preregistered runs or an explicitly budgeted mixture weight; they cannot be retroactively inserted into the old test.

The simple ratio assumes \(q_t\) is absolutely continuous with respect to \(p_t\). If the declared Born model assigns exact probability zero to an observed outcome, that observation already falsifies that exact null; in a real apparatus, calibration uncertainty should be incorporated into a conservative composite null rather than hidden by an arbitrary numerical floor.

At a conventional one-sided five-sigma tail level,

\[
\alpha=2.87\times10^{-7},\qquad
\frac1\alpha\approx3.484\times10^6,\qquad
\log E\approx15.064.
\]

Crossing that threshold is evidence against the **declared conditional null**. It is not by itself proof of a soul, extra spacetime, or a universal deterministic law.

### 6. Protocol P1 — prediction before the future setting

The strongest causal ordering is:

1. Prepare the target state and finish every declared control operation.
2. Record raw sensor/history data and compute \(\Lambda_t\).
3. Commit a cryptographic hash of the complete table \(q_t(y\mid s,\Lambda_t,C_t)\) for every allowed measurement setting \(s\).
4. Only then generate the random setting \(S_t\).
5. Perform the measurement and record \(Y_t\).
6. Reveal the committed prediction table and update \(E_T\).
7. Apply scheduled causal breaks, recalibrations, environmental resets, and process-tensor probes.
8. Repeat on a blinded holdout block and then on independently controlled hardware.

Committing a prediction for every setting before the setting is generated blocks ordinary setting leakage. Cryptographic commitment establishes ordering only when an independent timestamping or append-only service records the hash; a local hash created after the result proves nothing.

The first implementation should use Q1's three-qubit phase sector. Prepare operationally indistinguishable final states through histories with different canonical windings, then choose among incompatible measurement bases only after the prediction commitment. Standard Q1 phase coordinates and all tomography-supported memory enter \(p_t\). The residual winding lift competes only for information not already represented there.

If a signal appears, the mandatory elimination sequence is:

| Test | What survival would rule out |
|---|---|
| Randomized replay and label permutation | Coding, ordering, and analysis leakage |
| Sensor holdout and shielded dummy channels | Simple electromagnetic or timing side channels |
| Causal breaks and full reset | Ordinary apparatus memory |
| Process-tensor reconstruction | Detectable non-Markovian open-system dynamics; such dynamics can be reconstructed experimentally [Pollock et al. (2018)](https://link.aps.org/doi/10.1103/PhysRevA.97.012127) |
| State/measurement tomography | Misprepared state or miscalibrated POVM |
| Cross-platform replication | Device-specific drift |
| Spacelike Bell stage | Local predetermined-output explanations compatible with the test assumptions |
| No-signaling audit | Superluminal operational influence |

### 7. Power wall for a balanced binary benchmark

Suppose a predeclared predictor assigns probability \(a>1/2\) to its predicted bit and that bit is correct with long-run probability \(a\). Against \(p=(1/2,1/2)\), its mean log-evidence per trial is

\[
m(a)=a\log(2a)+(1-a)\log(2(1-a))
=D_{\mathrm{KL}}(\operatorname{Bern}(a)\Vert\operatorname{Bern}(1/2)).
\]

The following five-sigma planning counts use the exact mean for the expected crossing and a normal approximation for 90% power:

| Accuracy \(a\) | Advantage over chance | Expected-threshold trials | Approx. trials for 90% power |
|---:|---:|---:|---:|
| 55% | 5% | 3,008 | 4,773 |
| 52% | 2% | 18,825 | 29,899 |
| 51% | 1% | 75,314 | 119,635 |
| 50.5% | 0.5% | 301,271 | 478,579 |
| 50.1% | 0.1% | 7,531,887 | 11,964,788 |

These are planning approximations, not permission to stop at a favorable fixed-sample \(p\)-value. The e-process remains the primary test. Detector imbalance, dependence, drift, and a composite uncertain Born baseline require a conservative null family rather than the simple \(p=1/2\) benchmark.

### 8. Foundational walls that remain

Bell's theorem and loophole-free Bell experiments exclude the tested class of local hidden-variable explanations under their independence assumptions [Bell (1964)](https://link.aps.org/doi/10.1103/PhysicsPhysiqueFizika.1.195), [Hensen et al. (2015)](https://www.nature.com/articles/nature15759). Contextuality excludes a context-independent assignment of predetermined values to all observables. A viable \(\Lambda\) completion must therefore state whether it is contextual, nonlocal but operationally no-signaling, stochastic, or rejects a named independence assumption.

Colbeck and Renner prove, under their stated free-choice assumption and the correctness of quantum theory, that an extension cannot improve predictions of future measurement outcomes [Colbeck and Renner (2011)](https://www.nature.com/articles/ncomms1416). Their assumptions must be confronted explicitly; merely renaming \(\Lambda\) “dimensional” does not evade the theorem.

Bell violations are also used to certify randomness. Pironio et al. showed device-independent randomness generation from nonlocal correlations [Pironio et al. (2010)](https://arxiv.org/abs/0911.3427), and a later loophole-free experiment extracted 1,024 bits within \(10^{-12}\) of uniform under independence and no-superluminal-signaling assumptions [Bierhorst et al. (2018)](https://arxiv.org/abs/1803.06219). Therefore a predictor that genuinely penetrates device-independent certified bits would be a far stronger claim than predicting imperfections in a single qubit box.

### 9. What has and has not been broken

The **methodological wall is broken**: the missing variable, legal prediction rule, anti-cherry-picking statistic, causal ordering, rejection threshold, and power cost are now explicit and executable.

The **physical Born-randomness wall is not broken**. Breaking it requires a real \(\Lambda\) that crosses the preregistered e-value threshold on unseen trials, survives every apparatus-memory and leakage test, replicates independently, and remains compatible with all relevant Bell/contextuality/no-signaling data—or clearly identifies and tests the assumption it rejects.

This is the next frontier:

\[
\boxed{
h_t
\xrightarrow{\text{candidate lifts }L_k}
\Lambda_{t,k}
\xrightarrow{\text{Born tilt}}
q_{t,k}
\xrightarrow{\text{precommit + delayed setting}}
Y_t
\xrightarrow{\text{mixture e-process}}
\text{reject or kill the lift}.
}
\]

The executable companion born_outcome_frontier.py checks the martingale identity, the negative-KL wall, the mixture rule, cryptographic prediction commitments, the planning table, and synthetic null/positive worlds. Simulation validates the audit mechanism only; it is not empirical evidence for a dimensional predictor.

### 10. Composite-null hardening: uncertain calibration cannot be ignored

The simple e-process assumes that each \(p_t(y)\) is exactly known. A real experiment estimates \(\rho_t\), the POVM, detector response, leakage, and drift. Substituting point estimates and treating them as exact can create false evidence against Born's rule.

Let \(\mathcal P_t\) be a predictable set of conditional outcome distributions compatible with all calibration uncertainty retained under the null. Define the coordinatewise upper envelope

\[
u_t(y)=\sup_{p\in\mathcal P_t}p(y).
\]

For any normalized pre-outcome predictor \(q_t\), define

\[
e_t(Y_t)=\frac{q_t(Y_t)}{u_t(Y_t)},
\qquad
E_T^{\mathrm{rob}}=\prod_{t=1}^{T}e_t(Y_t).
\]

**Composite Born Envelope theorem.** If the actual conditional distribution \(p_t^*\) belongs to \(\mathcal P_t\), then

\[
\begin{aligned}
\mathbb E_{p_t^*}[e_t(Y_t)\mid\mathcal F_{t-1}]
&=\sum_y p_t^*(y)\frac{q_t(y)}{u_t(y)}\\
&\le\sum_y q_t(y)=1.
\end{aligned}
\]

Therefore \(E_T^{\mathrm{rob}}\) is a nonnegative supermartingale under every predictable sequence of null distributions in the declared family, and Ville's threshold remains valid. This construction is conservative because the separate coordinatewise suprema need not form one attainable distribution. That power loss is the price of refusing to convert calibration uncertainty into apparent new physics.

An exact-zero case needs explicit handling. If every \(p\in\mathcal P_t\) gives \(p(y)=0\), observing \(y\) falsifies that null family. If the zero came only from finite calibration or numerical truncation, the null family was incorrectly specified; inserting an arbitrary probability floor after seeing the outcome is forbidden.

Universal inference provides broader finite-sample tools for composite likelihood models [Wasserman, Ramdas, and Balakrishnan (2020)](https://www.pnas.org/doi/10.1073/pnas.1922664117). The envelope theorem above is a deliberately simple project-specific sufficient construction, not a claim to replace that literature.

### 11. Complexity-weighted lift search

RRLS generates candidate languages: phase unwraps, parity relations, winding registers, lagged histories, symmetry invariants, Walsh interactions, and boundary summaries. Testing every generated lift at full significance would make discovery inevitable under a large enough search.

Assign each frozen lift compiler \(L_k\) a binary prefix-code length \(\ell_k\) satisfying Kraft's inequality

\[
\sum_k2^{-\ell_k}\le1.
\]

Let \(E_{T,k}\) be any valid e-process for candidate \(k\). Define

\[
E_T^{K}
=1-\sum_k2^{-\ell_k}
+\sum_k2^{-\ell_k}E_{T,k}.
\]

**Kraft-weighted Search theorem.** \(E_T^K\) is a valid e-process because it is a fixed convex mixture of the constant process \(1\) and the candidate e-processes. A simple sufficient condition for candidate \(k\)'s weighted term alone to force \(E_T^K\ge1/\alpha\) is

\[
\boxed{
\log E_{T,k}
\ge
\log(1/\alpha)+\ell_k\log2.
}
\]

Each extra description bit costs one bit of evidence. This does not prove that shorter lifts are physically true, and it does not replace RRLS's polynomial recognition, representation-size, arithmetic-bit, and total-work obligations. It solves a different problem: statistical multiplicity.

At 51% accuracy in the balanced binary benchmark, charging code length changes the approximate five-sigma, 90%-power requirement as follows:

| Lift description length | Sufficient candidate \(\log E\) | Approx. trials |
|---:|---:|---:|
| 0 bits | 15.064 | 119,635 |
| 8 bits | 20.609 | 153,191 |
| 16 bits | 26.154 | 186,030 |
| 32 bits | 37.244 | 250,321 |

The code length and compiler grammar must be frozen before outcome scoring. Inventing a shorter encoding after a successful run is the statistical equivalent of changing a hypothesis after seeing the answer.

### 12. Projection-collision certificate and the RRLS connection

Let \(\Lambda\) be the full accessible pre-measurement record and let

\[
G=g(\Lambda,C)
\]

be a compressed dimensional address or invariant. Because \(G\) is a deterministic function of \((\Lambda,C)\), the conditional data-processing identity is

\[
I(Y;\Lambda\mid C)
=I(Y;G\mid C)
+I(Y;\Lambda\mid G,C).
\]

Hence

\[
\boxed{I(Y;G\mid C)\le I(Y;\Lambda\mid C).}
\]

Equality holds exactly when the discarded distinctions carry no further conditional outcome information. This yields a decisive collision certificate:

\[
G(\lambda,C)=G(\lambda',C)
\quad\text{but}\quad
P(Y\mid\lambda,C)\ne P(Y\mid\lambda',C)
\]

proves that no decoder receiving only \(G\) can be universally outcome-sufficient. Adding a more powerful classifier downstream cannot recover information already erased by the projection.

This is the same logical structure as the retained RRLS SAT/UNSAT invariant collision: equal compressed representation plus unequal required continuation kills that representation's sufficiency. The repair target is **boundary compatibility**, not collapse to zero. For quantum trials the boundary contains, at minimum, the prepared state, admissible measurement settings, timing, device identity, and allowed interventions. A valid canonicalizer must preserve the conditional relation from that boundary to future outcome distributions.

The executable audit checks a finite example in which a full lift contains \(0.192744757\) nats of outcome information and a coarsening that merges its two states contains exactly zero. This is an implementation check of data processing, not empirical quantum evidence.

### 13. Prediction is not yet causal or physical novelty

A successful BPO-2 predictor supports the statement “the declared standard context is predictively incomplete.” It does not identify the cause. The attribution ladder is:

| Gate | Required survival | What may be claimed |
|---|---|---|
| G0 | Held-out \(I(Y;\Lambda\mid C)>0\) and e-threshold crossing | Predictive residual |
| G1 | Raw-data audit, replay, negative controls, and immutable ordering | Not an analysis artifact |
| G2 | Tomography, detector model, process tensor, causal breaks, and resets | Not a modeled apparatus state or ordinary memory |
| G3 | Sign, magnitude, winding reversal, timing, and intervention invariance | Candidate coupling law rather than free curve fit |
| G4 | Independent platform and laboratory replication | Not device-specific |
| G5 | Bell/contextuality/free-setting/no-signaling compatibility | Candidate foundational extension |
| G6 | Lorentz behavior, classical limit, and independent magnitude derivation | Candidate physical theory |

Failure at a gate stops every stronger claim. In particular,

\[
I(Y;\Lambda\mid C)>0
\not\Rightarrow
\Lambda\text{ causes }Y
\not\Rightarrow
\Lambda\text{ is a new dimension}.
\]

If settings are chosen after the prediction commitment, the experiment must still test the setting-independence condition appropriate to its claim. In a Bell stage this includes a declared form of measurement independence. For locally accessible predictors, operational no-signaling requires that Alice's predicted marginal not depend on Bob's spacelike setting:

\[
P(A\mid X,Y,\Lambda_{\mathrm{local}})
=P(A\mid X,\Lambda_{\mathrm{local}}).
\]

A predictor that needs a future or remote setting unavailable at the prediction event has not predicted the local outcome in the claimed causal sense.

### 14. Hidden-variable cross-examination

Determinism alone does not imply accessible predictability. Bohmian mechanics is deterministic, yet its quantum-equilibrium analysis recovers \(\rho=|\psi|^2\) statistics and argues for an “absolute uncertainty” restricting subsystem configuration knowledge [Dürr, Goldstein, and Zanghì (1992)](https://link.springer.com/article/10.1007/BF01049004). Therefore “the universe may be deterministic” does not by itself break BPW.

The Colbeck–Renner conclusion depends on its explicit free-choice formulation. Ghirardi and Romano constructed predictively inequivalent ontological models under a different free-choice assumption [Ghirardi and Romano (2013)](https://link.aps.org/doi/10.1103/PhysRevLett.110.170404). This means the assumption boundary must be named, not that every predictive extension is viable.

Pilot-wave nonequilibrium has been proposed as a route to “subquantum information.” Valentini's signal-locality result shows the price: deterministic hidden-variable theories reproducing equilibrium quantum statistics generally expose instantaneous statistical signals for hypothetical nonequilibrium ensembles [Valentini (2002)](https://arxiv.org/abs/quant-ph/0112151). Any Dimensional Reality variant using that escape route inherits a direct no-signaling burden.

Finally, the likelihood-ratio strategy is not unprecedented in Bell analysis. Prediction-based-ratio methods already give memory-robust valid \(p\)-values for rejecting local realism even when states vary and local models use previous settings and outcomes [Zhang, Glancy, and Knill (2011)](https://link.aps.org/doi/10.1103/PhysRevA.84.062118). The possible contribution here is the integration of projection collisions, physically specified winding lifts, composite Born envelopes, and a frozen compiler grammar—not the invention of sequential likelihood ratios.

### 15. Frozen hardening protocol P2

The next implementation is frozen conceptually as follows:

1. Define the raw pre-measurement boundary record and forbid all downstream fields.
2. Freeze the Q1/Born composite null family \(\mathcal P_t\), including tomography and detector uncertainty.
3. Freeze a prefix-coded lift grammar before validation.
4. Generate \(\Lambda_{t,k}=L_k(h_t)\) using only admissible fields.
5. Fit each outcome tilt on past/training data only.
6. Commit all setting-conditional predictions before the delayed random setting.
7. Score with the composite-envelope factor and Kraft-weighted mixture.
8. Preserve negative results and exact projection collisions as permanent evidence against those lifts.
9. Spend one declared error budget across revisions; restarting software does not reset the investigation.
10. Advance any crossing through gates G1–G6 without changing its sign, magnitude law, code, or exclusions.

This protocol unifies the strongest existing disciplines in the project: DRPT projection closure, Q1 phase physics, BPW outcome scoring, the revision-1.26 conditional-filtration requirement, and RRLS boundary-compatible canonicalization. It remains a protocol awaiting real quantum records.

The separate verify_born_outcome_frontier.py imports none of the production module. It independently reconstructs the fixed examples, 5,000 random composite-null and data-processing checks, both 120,000-trial synthetic paths, and the code-length power table. Agreement is implementation independence within the same formulas and pseudorandom generator, not external replication or physical evidence.

### 16. Information rate and Wallbreaker Protocol P3

Let

\[
r_t(y)=P(Y_t=y\mid C_t,\Lambda_t)
\]

be the true outcome law conditional on the complete standard context and an admissible pre-measurement lift. Write \(\bar r_t(y)=P(Y_t=y\mid C_t)\) for its actual marginal and \(b_t(y)\) for the declared Born baseline. For any candidate decoder \(q_t\), direct expansion of the logarithm gives

\[
\boxed{
\mathbb E\log\frac{q_t(Y_t\mid C_t,\Lambda_t)}{b_t(Y_t\mid C_t)}
=I(Y_t;\Lambda_t\mid C_t)
+\mathbb E_{C_t}D_{\mathrm{KL}}(\bar r_t\Vert b_t)
-\mathbb E D_{\mathrm{KL}}(r_t\Vert q_t).
}
\]

This is the **general Born-evidence decomposition**. The second term is marginal Born-rule violation and must not be mislabeled as hidden individual-outcome information. For a Born-marginal-preserving extension, \(\bar r_t=b_t\), that term vanishes and the **Information-Rate identity** remains:

\[
\mathbb E\log\frac{q_t(Y_t\mid C_t,\Lambda_t)}{b_t(Y_t\mid C_t)}
=I(Y_t;\Lambda_t\mid C_t)
-\mathbb E D_{\mathrm{KL}}(r_t\Vert q_t)
\le I(Y_t;\Lambda_t\mid C_t).
\]

The largest possible expected conditional-prediction rate in the marginal-preserving case is exactly \(I(Y;\Lambda\mid C)\), attained only by the true extended conditional law. More computation can reduce decoder misspecification but cannot create information missing from the admissible boundary record.

Under stationary independent trials and an exact decoder, \(T^{-1}\log E_T\to I(Y;\Lambda\mid C)\) almost surely. A prefix-coded lift of length \(\ell\) therefore has the characteristic planning scale

\[
T_{\rm scale}\approx
\frac{\log(1/\alpha)+\ell\log2}{I(Y;\Lambda\mid C)}.
\]

This is an asymptotic scale, not an exact stopping-time expectation. Positive conditional information also does not imply deterministic outcomes; exact individual prediction requires \(H(Y\mid C,\Lambda)=0\).

No finite experiment can prove an unrestricted continuation law. Given any proper finite tested set \(S\subset\mathcal C\), a second law can equal the first on \(S\) and differ at an untested context. The executable collision makes the limitation concrete: two laws have total-variation distance zero on seven tested contexts and distance \(0.8\) on the eighth. “Universal” therefore requires a restricted model class with a proved extension rule, followed by sealed context expansion.

Wallbreaker Protocol P3 scores three different evidence streams:

\[
E_T^{\rm outcome},\qquad
E_T^{\rm setting},\qquad
E_T^{\rm signal}.
\]

The first searches for outcome information beyond the complete calibrated context. The second detects whether a proposed lift predicts the delayed supposedly free setting. The third detects dependence of a local outcome on a remote setting. The desired signature is an outcome crossing without setting or signaling evidence, together with structural isolation and affirmative equivalence controls; failure to reject setting dependence or signaling does not prove their exact absence.

Six seeded 50,000-trial adversarial worlds separate the routes:

| Synthetic world | Complete-context outcome crossing | Setting crossing | Signaling crossing | Verdict |
|---|---:|---:|---:|---|
| Born null | No | No | No | Correct negative control |
| Planted accessible lift | Yes | No | No | Target statistical signature; synthetic only |
| Future-setting leak | Yes | Yes | Yes | Disqualified by chronology and measurement independence |
| Remote signaling | No | Yes | Yes | Naive outcome evidence disappears when the remote setting enters \(C\) |
| Apparatus memory | No | No | No | Naive evidence disappears when ordinary memory enters \(C\) |
| Memory after reset | No | No | No | Reset kills the apparent predictor |

Wallbreaker Audit 0.1.0 and an implementation that imports none of its production functions agree on all six classifications. They also reproduce the marginal-preserving information rate \(0.368064207168\) nats, the nonpreserving decomposition \(0.388475204429=0.368064207168+0.0204109972601\) nats, and 5,000 randomized decomposition checks with maximum identity errors \(1.33\times10^{-15}\) in production and \(1.11\times10^{-15}\) independently. This is finite synthetic evidence about the audit, not physical evidence against Born's rule.

The foundational escape route must be declared before data. Kochen and Specker exclude the relevant noncontextual value assignments, not every contextual hidden-variable model [Kochen and Specker (1967)](https://www.jstor.org/stable/24902153). Bell-type conclusions require an explicit measurement-independence premise; Hall gives a deterministic no-signaling singlet model after quantifiably relaxing that premise [Hall (2010)](https://link.aps.org/doi/10.1103/PhysRevLett.105.250404). This demonstrates assumption sensitivity, not a cost-free physical completion.

P3 therefore retains thirteen separate walls: internal typing, Born information, projection collisions, chronology, measurement independence, context completeness, no-signaling, contextuality, the Bell route, Lorentz behavior and scale, finite-context extension, computational complexity, and external replication. The complete pass/fail ledger and required immutable trial schema are preserved in `Wallbreaker_Protocol_P3.md`. W1–W9 and W12 retain physical obligations, with different parts already resolved only at the mathematical or protocol level.

### 17. P3-R1 hardware package: admissibility ruling

The hardware frontier is no longer only a verbal next step. P3-R1 fixes a three-qubit equal-magnitude eight-phase state, `Z_ALL` and `X_ALL` settings, one-shot circuit/PUB records, three prefix-coded winding probes, a composite calibration envelope, a prediction hash chain, a delayed setting ledger, a setting-leak e-process, crash-safe provider-job recovery, and an independent verifier.

For each frozen candidate, the probability added to one outcome is removed from its \(\oplus4\) partner. Across all eight windings the reverse transfer appears once with the same magnitude, proving exact cycle-marginal preservation. The construction can therefore search for conditional outcome information without inserting a cycle-averaged marginal violation into its ideal reference model. It remains a test coordinate, not a physical derivation of an eighth dimension.

At 20,000 synthetic trials, the exact Born-reference control crosses neither threshold; a planted lift crosses the outcome threshold but not the setting threshold; a contaminated lift crosses both and is disqualified. The maximum outcome/setting log mixtures are respectively \((0.279025,4.631600)\), \((25.026590,1.346171)\), and \((18.835820,1094.733929)\). Independent replay agrees, the run-specific cycle-marginal numerical error is \(5.55\times10^{-17}\), and 5,000 randomized checks remain within floating-point tolerance. Altering one commitment causes a digest failure.

Admissibility limitations are explicit:

1. No QPU record exists. The distributed config refuses live execution without a fresh UUID, named accessible backend, and bound calibration artifact.
2. The 0.02 per-outcome future-drift margin is a predeclared model allowance, not a theorem about hardware stability.
3. Local SHA-256 manifests detect changes relative to themselves but need an external timestamp/public anchor to establish third-party chronology.
4. IBM Runtime returns job submission and result retrieval information, not an authenticated device timestamp for each shot.
5. One processor is not spacelike separated. W6 no-signaling, W7 contextuality, and W8 Bell claims receive no pass or fail from P3-R1.
6. A 256-trial pilot has not been powered for small new-physics effects. Its outcome cannot justify an unrestricted or universal conclusion.

The court-grade ruling is therefore: **the measurement instrument is executable and adversarially tested; the physical exhibit is absent**. The next admissible act is to activate one explicit backend, obtain independent calibration, externally anchor the frozen manifest, execute without edits, and replay the raw provider records. Anything stronger would confuse readiness with evidence.

---

## XVII. Existing empirical pressure on extra dimensions

Extra coordinates are not ruled out in every conceivable form, but physical extra spatial dimensions usually change forces, spectra, or propagation unless compactified or decoupled. A torsion-balance experiment found Newtonian gravity to fit data down to 52 micrometers and constrained gravitational-strength Yukawa deviations [Lee et al. (2020)](https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.124.101101).

This does not directly refute the present eight-cycle model because its geometry, scale, and coupling are unspecified. That is also the problem: without those quantities, the “eight dimensions” claim evades both confirmation and refutation.

---

## XVIII. Cross-examination: likely questions and honest answers

### “Where does the number eight come from?”

Eight now has an exact computational interpretation: it is the number of computational basis states of three qubits. Eight raw basis phases reduce to seven relative phases after common phase is removed. What remains unproved is why this three-qubit sector should be a fundamental description of physical reality rather than one useful model among many.

### “What does the theory predict that quantum mechanics does not?”

The Q1 phase compiler, alignment detector, cube walk, and entanglement equations reproduce standard quantum mechanics. B1 proposes a winding-dependent Ramsey shift and B2 proposes an irreducible eight-path term. Those are the only current beyond-standard predictions, and both remain provisional.

### “Have we built the mathematics for a quantum computer?”

We have built and verified a clean three-qubit diagonal phase compiler and added the operations needed to place it inside a universal circuit language. That is a legitimate quantum-computation architecture. It is not new quantum mechanics, physical hardware, or evidence of computational advantage; scalability, sparse problem structure, error correction, and physical control remain separate obligations.

### “Can the framework now predict an individual Born-random outcome?”

No empirical predictor has been found. The framework now supplies the precise search and falsification method: an accessible pre-measurement lift must earn held-out log evidence against the complete Born conditional model. A synthetic predictor crosses the threshold in simulation, showing that the test detects real information when present; that does not show nature supplies such information.

### “Does consciousness collapse the wavefunction?”

No such result has been derived. Physical which-path records and environmental decoherence already suppress interference without human awareness. A soul-collapse law would need its own probability rule.

### “Have Bell tests ruled out your hidden dimensions?”

Not all hidden-variable theories. They rule out the tested class of local explanations satisfying the relevant independence assumptions. The model must declare whether it is contextual, nonlocal but no-signaling, stochastic, or rejects another assumption.

### “Is ‘all potential at zero’ mathematics or poetry?”

It becomes mathematics only as an admissible set \(\mathcal P_0\) or measure \(\mu_0\) on a defined state space. The vector zero alone is insufficient.

### “Why can time never align again?”

It can align again when frequencies are commensurate. Exact nonrecurrence requires an incommensurability condition. Approximate recurrence may still occur.

### “Does the model explain the double slit?”

It can reproduce the standard overlap/decoherence calculation. That is consistency, not a new solution. A new solution requires a distinct, tested prediction.

### “Does observation expand space?”

Accumulating records expands the known/experienced domain and shrinks compatible-history fibers. No equation yet links that process to cosmological metric expansion.

### “What is the strongest result today?”

The eight-phase lift is exactly invertible, its common phase leaves seven observable Pauli-interaction coordinates, computational-basis probabilities have closure defect one under allowed Hadamard intervention, and the phase geometry gives exact output and entanglement predictions. Separately, the higher-order-null theorem turns an all-eight nonstandard coupling into an \(I_8\) test. BPW/P2 now makes individual-outcome searches safe against a declared composite Born null and charges compiler-generated lifts for multiplicity. None of these results establishes that nature contains B1, B2, or an accessible outcome variable.

### “What would change your mind?”

Failure to derive a non-arbitrary bridge law, experimental bounds forcing its coupling to zero, violation of no-signaling, incompatibility with Bell/contextuality data, or inability to recover established limiting physics would count against the physical theory.

---

## XIX. Evidentiary standard going forward

Every statement should carry one label:

- **D — Definition:** true because terminology was chosen that way.
- **T — Theorem:** follows deductively from stated assumptions.
- **E — Empirical result:** supported by cited measurements.
- **M — Model correspondence:** reproduces an established equation or behavior.
- **H — Hypothesis:** could be tested but is not established.
- **A — Axiom/bridge:** added assumption connecting mathematical layers.
- **P — Philosophical/metaphysical proposition:** meaningful but not operationally tested as stated.
- **R — Retracted/corrected claim:** no longer asserted.

The framework should never use a D, M, finite simulation, or metaphor as if it were E.

---

## XX. Bottom line

The theory is not unrefutable and should not try to be. Its defensible skeleton is:

\[
\boxed{
\text{hidden state}
\xrightarrow{\text{observation }\pi}
\text{record}
\xrightarrow{\text{closure test}}
\text{missing predictive state}
\xrightarrow{\text{bridge law}}
\text{new measurable prediction}
}
\]

The first three arrows are mathematically respectable and connect to established fields. Q1 now supplies an exact standard-quantum implementation of the eight-phase geometry: compiler, alignment measurement, entanglement prediction, and cube movement. BPW/P2 supplies an exact individual-outcome frontier: pre-measurement lifts, a composite Born envelope, prefix-coded search penalties, prediction commitments, delayed settings, and a valid e-process. P3 now quantifies the maximum evidence rate and separates a candidate outcome signal from setting leakage, signaling, and memory. The last arrow remains the unsolved new-physics problem. B1 tests absolute winding memory; B2 tests an irreducible eight-path alignment interaction; BPW/P3 tests whether any accessible lift predicts individual outcomes beyond the complete Born conditional model. All remain provisional. Only a non-arbitrary coupling and a successful preregistered experiment can turn geometry into new physics.

That is the position that can survive a hostile room: admit what is known, display the proofs, identify the prior art, state the assumptions, and expose the physical hypothesis to a test capable of killing it.

## Primary-source reference set

1. B.-G. Englert, “Fringe Visibility and Which-Way Information: An Inequality,” *Physical Review Letters* 77, 2154 (1996): https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.77.2154
2. S. Dürr, T. Nonn, and G. Rempe, “Fringe Visibility and Which-Way Information in an Atom Interferometer,” *Physical Review Letters* 81, 5705 (1998): https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.81.5705
3. M. Brune et al., “Observing the Progressive Decoherence of the ‘Meter’ in a Quantum Measurement,” *Physical Review Letters* 77, 4887 (1996): https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.77.4887
4. B. Hensen et al., “Loophole-free Bell inequality violation using electron spins separated by 1.3 kilometres,” *Nature* 526, 682–686 (2015): https://www.nature.com/articles/nature15759
5. G. Kirchmair et al., “State-independent experimental test of quantum contextuality,” *Nature* 460, 494–497 (2009): https://arxiv.org/abs/0904.1655
6. M. F. Pusey, J. Barrett, and T. Rudolph, “On the reality of the quantum state,” *Nature Physics* 8, 475–478 (2012): https://arxiv.org/abs/1111.3328
7. C. Simon, V. Bužek, and N. Gisin, “No-Signaling Condition and Quantum Dynamics,” *Physical Review Letters* 87, 170405 (2001): https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.87.170405
8. J.-D. Bancal et al., “Quantum nonlocality based on finite-speed causal influences leads to superluminal signaling,” *Nature Physics* 8, 867–870 (2012): https://arxiv.org/abs/1110.3795
9. M. Proietti et al., “Experimental test of local observer independence,” *Science Advances* 5, eaaw9832 (2019): https://arxiv.org/abs/1902.05080
10. J. P. Crutchfield and K. Young, “Inferring statistical complexity,” *Physical Review Letters* 63, 105 (1989): https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.63.105
11. M. L. Littman and R. S. Sutton, “Predictive Representations of State,” *NeurIPS 14* (2001): https://proceedings.neurips.cc/paper/2001/hash/1e4d36177d71bbb3558e43af9577d70e-Abstract.html
12. M. N. Jacobi and O. Görnerup, “A dual eigenvector condition for strong lumpability of Markov chains,” arXiv:0710.1986 (2008): https://arxiv.org/abs/0710.1986
13. J. G. Lee et al., “New Test of the Gravitational Law at Separations down to 52 μm,” *Physical Review Letters* 124, 101101 (2020): https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.124.101101
14. A. Tonomura et al., “Evidence for Aharonov–Bohm Effect with Magnetic Field Completely Shielded from Electron Wave,” *Physical Review Letters* 56, 792 (1986): https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.56.792
15. P. J. Leek et al., “Observation of Berry’s Phase in a Solid-State Qubit,” *Science* 318, 1889–1892 (2007): https://arxiv.org/abs/0711.0218
16. K. Saito and T. Mukaiyama, “Sensing Aharonov–Bohm Phase Using a Multiply-Orbiting-Ion Interferometer,” *Physical Review Letters* 133, 143402 (2024): https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.133.143402
17. Y.-H. Kim et al., “A Delayed Choice Quantum Eraser,” *Physical Review Letters* 84, 1 (2000): https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.84.1
18. F. A. Pollock et al., “Operational Markov Condition for Quantum Processes,” *Physical Review Letters* 120, 040405 (2018): https://arxiv.org/abs/1801.09811
19. G. A. L. White et al., “Demonstration of non-Markovian process characterisation and control on a quantum processor,” *Nature Communications* 11, 6301 (2020): https://www.nature.com/articles/s41467-020-20113-3
20. R. Wagner et al., “Violation of a Leggett–Garg Inequality Using Ideal Negative Measurements in Neutron Interferometry,” *Physical Review Letters* 132, 260201 (2024): https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.132.260201
21. S. Pedalino et al., “Probing quantum mechanics with nanoparticle matter-wave interferometry,” *Nature* 649, 866–870 (2026): https://www.nature.com/articles/s41586-025-09917-9
22. R. D. Sorkin, “Quantum Mechanics as Quantum Measure Theory,” *Modern Physics Letters A* 9, 3119–3127 (1994): https://arxiv.org/abs/gr-qc/9401003
23. U. Sinha et al., “Ruling Out Multi-Order Interference in Quantum Mechanics,” *Science* 329, 418–421 (2010): https://arxiv.org/abs/1007.4193
24. F. Jin et al., “Experimental Test of Born’s Rule by Inspecting Third-Order Quantum Interference on a Single Spin in Solids,” *Physical Review A* 95, 012107 (2017): https://arxiv.org/abs/1612.08563
25. X. Chen et al., “A generalized multipath delayed-choice experiment on a large-scale quantum nanophotonic chip,” *Nature Communications* 12, 2712 (2021): https://www.nature.com/articles/s41467-021-22887-6
26. U. Sinha et al., “Nonclassical Paths in Quantum Interference Experiments,” *Nature Communications* 8, 13987 (2017): https://www.nature.com/articles/ncomms13987
27. P. Namdar et al., “Experimental Higher-Order Interference in a Nonlinear Triple Slit,” *Physical Review A* 107, 032211 (2023): https://journals.aps.org/pra/abstract/10.1103/PhysRevA.107.032211
28. F. Bouchard et al., “Programmable Photonic Quantum Circuits with Ultrafast Time-Bin Encoding,” *Physical Review Letters* 133, 090601 (2024): https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.133.090601
29. L. Carosini et al., “Programmable multiphoton quantum interference in a single spatial mode,” *Science Advances* 10, eadj0993 (2024): https://www.science.org/doi/10.1126/sciadv.adj0993
30. J. Welch et al., “Efficient Quantum Circuits for Diagonal Unitaries Without Ancillas,” *New Journal of Physics* 16, 033040 (2014): https://arxiv.org/abs/1306.3991
31. M. Rossi et al., “Quantum Hypergraph States,” *New Journal of Physics* 15, 113022 (2013): https://arxiv.org/abs/1211.5554
32. D. Deutsch and R. Jozsa, “Rapid solution of problems by quantum computation,” *Proceedings of the Royal Society A* 439, 553–558 (1992): https://royalsocietypublishing.org/doi/10.1098/rspa.1992.0167
33. E. Bernstein and U. Vazirani, “Quantum Complexity Theory,” *SIAM Journal on Computing* 26, 1411–1473 (1997): https://epubs.siam.org/doi/10.1137/S0097539796300921
34. A. Barenco et al., “Elementary gates for quantum computation,” *Physical Review A* 52, 3457–3467 (1995): https://arxiv.org/abs/quant-ph/9503016
35. P. O. Boykin et al., “On Universal and Fault-Tolerant Quantum Computing,” *Information Processing Letters* 75, 101–107 (2000): https://arxiv.org/abs/quant-ph/9906054
36. D. Aharonov and M. Ben-Or, “Fault-Tolerant Quantum Computation With Constant Error Rate,” arXiv:quant-ph/9906129 (1999): https://arxiv.org/abs/quant-ph/9906129
37. J. S. Bell, “On the Einstein Podolsky Rosen paradox,” *Physics Physique Fizika* 1, 195–200 (1964): https://link.aps.org/doi/10.1103/PhysicsPhysiqueFizika.1.195
38. R. Colbeck and R. Renner, “No extension of quantum theory can have improved predictive power,” *Nature Communications* 2, 411 (2011): https://www.nature.com/articles/ncomms1416
39. S. Pironio et al., “Random Numbers Certified by Bell's Theorem,” *Nature* 464, 1021–1024 (2010): https://arxiv.org/abs/0911.3427
40. P. Bierhorst et al., “Experimentally Generated Randomness Certified by the Impossibility of Superluminal Signals,” *Nature* 556, 223–226 (2018): https://arxiv.org/abs/1803.06219
41. P. Grünwald, R. de Heide, and W. Koolen, “Safe Testing,” *Journal of the Royal Statistical Society Series B* 86, 1091–1128 (2024): https://academic.oup.com/jrsssb/article/86/5/1091/7617024
42. S. R. Howard, A. Ramdas, J. McAuliffe, and J. Sekhon, “Time-uniform Chernoff bounds via nonnegative supermartingales,” *Probability Surveys* 17, 257–317 (2020): https://arxiv.org/abs/1808.03204
43. L. Wasserman, A. Ramdas, and S. Balakrishnan, “Universal inference,” *Proceedings of the National Academy of Sciences* 117, 16880–16890 (2020): https://www.pnas.org/doi/10.1073/pnas.1922664117
44. Y. Zhang, S. Glancy, and E. Knill, “Asymptotically optimal data analysis for rejecting local realism,” *Physical Review A* 84, 062118 (2011): https://link.aps.org/doi/10.1103/PhysRevA.84.062118
45. D. Dürr, S. Goldstein, and N. Zanghì, “Quantum equilibrium and the origin of absolute uncertainty,” *Journal of Statistical Physics* 67, 843–907 (1992): https://link.springer.com/article/10.1007/BF01049004
46. G. C. Ghirardi and R. Romano, “Ontological Models Predictively Inequivalent to Quantum Theory,” *Physical Review Letters* 110, 170404 (2013): https://link.aps.org/doi/10.1103/PhysRevLett.110.170404
47. A. Valentini, “Signal-Locality and Subquantum Information in Deterministic Hidden-Variables Theories,” in *Non-Locality and Modality* (2002): https://arxiv.org/abs/quant-ph/0112151
48. A. Barron, J. Rissanen, and B. Yu, “The Minimum Description Length Principle in Coding and Modeling,” *IEEE Transactions on Information Theory* 44, 2743–2760 (1998): https://web.mit.edu/6.454/www/www_fall_2003/barron.pdf
49. M. J. W. Hall, “Local Deterministic Model of Singlet State Correlations Based on Relaxing Measurement Independence,” *Physical Review Letters* 105, 250404 (2010): https://link.aps.org/doi/10.1103/PhysRevLett.105.250404
50. S. Kochen and E. P. Specker, “The Problem of Hidden Variables in Quantum Mechanics,” *Journal of Mathematics and Mechanics* 17, 59–87 (1967): https://www.jstor.org/stable/24902153