1. The experiment we actually want to run#

Executive thesis#

The central experiment is deliberately simple. The physics program has already committed a typed Shape: a Stage, a Rulebook, Actors, Co-Actors, invariant class relations, and a full-precision 13D instance. Instead of tuning that Shape to a quantum-computing benchmark, this dossier freezes it and asks what happens when a compiler is required to consume every component while constructing a fault-tolerant logical memory and its supporting control architecture.

The proposed evidence is not that the Shape uniquely forces a code. Uniqueness is unnecessary. A bridge beam is load-bearing if removing it changes what the structure can carry after the remaining structure has been allowed to settle. The analogous test here is ablate -> legally re-optimize -> measure -> restore. A component earns a load-bearing label only if its removal causes the preregistered failure signature, the loss survives legal re-optimization, and restoring the original component restores the capability. If a substitute geometry recovers the capability, the component is load-bearing but not unique. If nothing changes, it is redundant at the tested scope.

The existing source corpus is especially useful because it already contains a no-feedback firewall: the quantum-computing numbers are external/simulated and may not feed back into GUT gate statuses or geometry selection. That separation is not an inconvenience; it is what makes a prospective cross-domain test interpretable. The engineering bridge cannot have been used to choose the physics answer if the freeze and hash chronology are respected.

Source freeze and reproducibility#

2. Quantum-error-correction target and common mathematical language#

1.1 Minimal quantum-error-correction language#

A quantum error-correcting code is a protected subspace (or subsystem) of a larger physical Hilbert space. In stabilizer language, commuting observables define the code by a simultaneous eigenspace. If the stabilizer generators are S_i, the ideal projector onto the +1 code sector is

This equation is the most direct external analogue of the existing engineering bridge in the GUT/Quantum corpus: admit one eigenspace, reject the rest. It is standard stabilizer-code mathematics, not a new physics claim. The bridge is interesting only because the same operation appears upstream as an admissibility/chamber operation on the frozen Shape.

For a correctable error set {E_a}, the Knill-Laflamme condition can be written

The design problem is therefore naturally a typed one: define a support, define allowed operators, define the protected sector, define the error complement, and define a recovery map. Those five jobs align closely with Stage, Rulebook, Actors, Co-Actors, and Dynamics without requiring the claim that the two theories are identical.

1.2 Exact micro-certificate: the projector really is an admissibility operation#

As a sanity check, consider the three-qubit repetition-code stabilizers S_1=Z_1Z_2 and S_2=Z_2Z_3. The projector

has P^2=P exactly and trace(P)=2, so it admits a two-dimensional code sector and annihilates the orthogonal syndrome sectors. The numerical check used while assembling this dossier returned an idempotence residual of 0.0 at machine precision and trace 2.0. This is not evidence for the 13D geometry; it is the exact coding operation that makes the chamber translation mathematically concrete.

1.3 Target engineering gate#

The worked engineering gate used throughout the dossier is QC-G1: protected logical memory under realistic error channels with minimum total implementation cost. The gate is intentionally narrower than "build a universal quantum computer." A candidate must specify the protected subspace, syndrome/error complement, allowed couplings, boundary/control interfaces, recovery/decoder interface, and a resource ledger. Later logical gates are treated as a downstream extension.

3. Compiler architecture: how every Shape component is allowed to matter#

2.1 The Shape-to-QEC compiler contract#

The compiler is not allowed to treat "13D" as a magic scalar. It must consume the typed content. Its output is a code/architecture tuple

where H_phys is the physical Hilbert space, S the stabilizer or constraint algebra, P_C the protected-space projector, E the error set, R the recovery interface, G the interaction/connectivity graph, B the boundary/control interface, and D the decoder/dynamics contract. Every output field must name which Shape rows constrained it.

2.2 Nature-derived tensors enter after the Shape, not instead of it#

The user-defined design workflow adds another independent source of constraint. For each engineering challenge, identify several natural systems that solve an analogous functional problem, abstract the common mechanism, represent it as a tensor/constraint, and search only inside the reduced admissible space. The tensor is not allowed to overwrite Shape; it is attached to the relevant Stage/Rulebook/Actor slots.

Examples include a locality tensor L_ij for wiring cost, an error-covariance tensor C_ij for correlated noise, a propagation tensor G^k_ij for fault spread, a redundancy/incidence tensor B for alternate recovery paths, and a projector P_adm for admissible states or architectures. These are engineering constraints. Their role is to reduce the search space, while the Shape supplies the structural coordinate system in which they are placed.

4. Observer spacetime M4 - causal support, not a decorative prefactor#

Physics function#

Shape V8.1 defines M4 as the four-dimensional Lorentzian comparison surface with dynamic metric g_mu_nu(x). It owns local causal support and the dimensional surface on which records are compared; it does not own the dynamics or the measurement map. In the physics program this prevents raw internal or 13D objects from being compared directly with a four-dimensional observable without a declared projection.

Worked physics example#

The SG-8 history recorded in the project assumptions ledger is the canonical same-ruler example: a raw 13D flavor amplitude was initially compared directly with a one-chamber 4D running mass. The repair was not to dissolve the measured quark mass; it was to supply the missing 13D-to-4D projection coefficient. This shows why observer support is a real structural field rather than prose.

Quantum-design consumption#

In QC-G1, M4 is consumed as the rule that hardware operations must admit a causal scheduling and locality interpretation. The compiler may use nonlocal graph edges only if the hardware model explicitly pays for transport/coupler resources and timing. The Shape does not forbid abstract nonlocal codes; it forbids hiding the physical realization cost.

Load-bearing intervention#

Required cross-layer dependencies#

• Rulebook R-15/R-16 for ruler and observer projection
• Actor AC-11/AC-12 for response and causality
• Scale for physical communication/time costs

5. K6 = SU(3)/T2 - the strongest direct geometric bridge#

Physics function#

K6 is the six-dimensional complete flag manifold SU(3)/T^2. In the current physics construction it supplies the color/family carrier, the A2 root-plane tangent decomposition, the Weyl structure, a nontrivial cohomology shelf, and the spin-C family-index stage. Shape V8.1 freezes the isotropic metric point (u1,u2,u3)=(1,1,1), radius R6=R0, Euler characteristic 6, and exact normalized curvature invariants.

Worked physics example#

The term dossier records that removing K6 simultaneously removes the internal stage for the SU(3) color carrier and detaches the topological family-count mechanism. The relevant line-bundle/spin-C index is recorded as -3 in the current GUT certificate chain. Importantly, the dossier does not claim K6 alone closes the Standard Model; it names the bundle and Rulebook dependencies separately.

Quantum-design consumption#

The existing engineering bridge already names K6 as part of the QC translation, so this component is not introduced after the fact for this dossier. The compiler uses three distinct K6 structures: (i) the six Weyl chambers/S3 action to define symmetry-equivalent routing sectors; (ii) the three A2 root planes to define three sparse interaction families; and (iii) the nontrivial topology/symmetry to restrict candidate code cells before numerical search.

Load-bearing intervention#

Required cross-layer dependencies#

• C_admiss for chamber admission
• Actor assignments for physical carriers
• Rigidity for deformation tests
• Interdependence for downstream replay

6. S2 - weak-sector sphere as a paired representation/routing primitive#

Physics function#

The round two-sphere is the two-dimensional weak-sector carrier. Its rotational symmetry lifts to SU(2) on spinors; the Cartan generator supplies T3, and the bundle/monopole-sector structure distinguishes singlets from doublets. The term dossier makes the layer boundary explicit: the manifold is Stage, while the principal SU(2)L bundle and representation modules are Actor/tensor data.

Worked physics example#

The worked Standard-Model example is the doublet/singlet routing. Q_L, L_L and the Higgs occupy weak-doublet sectors, while u_R, d_R and e_R are singlets. The mixed SU(2)^2-U(1) anomaly trace vanishes on one generation because 3(1/6)-1/2=0. The calculation does not prove the geometry, but it demonstrates that the sphere-associated representation structure is used quantitatively downstream.

Quantum-design consumption#

The QC compiler consumes the SU(2)-type structure as a paired complementary-channel primitive: a protected local cell has a two-component syndrome/control representation, while the three generators specify rotations among equivalent local bases. This is useful when an error mechanism is basis-dependent: the design is scored on performance across the SU(2)-related basis family rather than one hand-picked axis.

Load-bearing intervention#

Required cross-layer dependencies#

• Gauge/control Actor owns the actual physical operations
• Co-Actor error model must include basis-biased noise
• Dynamics defines allowed rotations/schedules

7. Parent circle S1Y - periodic phase and winding information#

Physics function#

S1Y is the parent hypercharge cover. It carries a closed one-cycle before the Z2 quotient and is the natural home for periodic phase/winding data. Shape V8.1 carefully separates the cover from the active interval: importing a closed-cycle result into the quotient without parity and normalization conversion is forbidden.

Worked physics example#

In the physics construction, the circle/cover structure is tied to the hypercharge presentation and to Wilson/holonomy data used elsewhere. The exact role depends on whether a statement is cover-level or quotient-level; the current Stage file explicitly forbids silently treating the active interval as if it retained the parent closed cycle.

Quantum-design consumption#

The QC compiler uses the parent-circle information only for operations that genuinely require a periodic control coordinate: phase-space cycles, cyclic syndrome schedules, or holonomic control loops. This is deliberately separated from the quotient/parity filter in the next chapter.

Load-bearing intervention#

Required cross-layer dependencies#

• IY quotient owns parity/fixed boundaries
• Higgs/Wilson Actor owns the protected phase operation
• Scale owns the physical period/time/energy conversion

8. IY = S1Y/Z2 and fixed strata - projection as a physical design operation#

Physics function#

The active orbifold interval is obtained by the reflection theta -> -theta. It has two fixed endpoints and no active closed one-cycle. In the physics chain, parity and boundary-domain data remove unwanted mirror modes and participate in the chirality certificate. This is one of the cleanest "geometry does work" examples because the quotient changes the allowed spectrum rather than merely renaming coordinates.

Worked physics example#

The current chirality narrative combines the K6 index with the interval projection: the index counts three family copies and the boundary/parity projection removes the mirror. The two fixed strata F0 and Fpi carry opposite normal orientations and are therefore not interchangeable when operator domains or inflow terms are evaluated.

Quantum-design consumption#

In QC-G1, the quotient becomes an explicit leakage/error-sector projection rule. A parity operator Pi splits the local Hilbert space; the code admits the desired parity sector and treats the complement as a Co-Actor. Fixed strata become privileged syndrome/reset/control interfaces rather than additional data sites.

Load-bearing intervention#

Required cross-layer dependencies#

• Boundary/domain Rulebook R-05 and chirality rule R-08
• Co-Actor CA-DOMAIN/CA-CHIRALITY for the rejected sector
• Dynamics for physical leakage and reset channels

9. Product Stage X9/X13 and block metric - modular ownership without hidden cross-couplings#

Physics function#

The active internal Stage is X9=K6 x S2 x IY and the complete Stage is X13=M4 x X9. The frozen metric is block-product; undeclared external/internal mixed metric components and undeclared off-diagonal internal terms are outside the current physical Stage. This is a strong ownership statement: different structural jobs have distinct carriers, and cross-couplings must be declared rather than appearing implicitly.

Worked physics example#

In the physics model, this factorization keeps the color/family, weak, and hypercharge/chirality jobs distinct while still allowing them to meet through the Rulebook and bundle Actors. The term dossiers explicitly reject the idea that one carrier can be silently used as another.

Quantum-design consumption#

The QEC compiler uses the same principle as modular architectural factorization: inner bosonic protection, finite chamber selection, topological outer protection, and modular scaling are typed as different modules with explicit interfaces. A coupling between modules is charged as an interface Actor rather than hidden inside one module's local cost.

Load-bearing intervention#

Required cross-layer dependencies#

• Actor one-owner rule AC-03/AC-13
• Rulebook R-18 canonical one-owner rule
• Interdependence to replay interface changes

10. A2 root-plane decomposition - three structured coupling families#

Physics function#

At the homogeneous level, the tangent space of K6 decomposes into three real two-planes associated with the three positive A2 roots. Shape V8.1 identifies this as the complete homogeneous metric-extension basis and uses it in the rigidity calculation. It is therefore more detailed than merely saying "there is an SU(3) symmetry."

Worked physics example#

The physics use is the allowed homogeneous deformation/coupling basis: changing the three root-plane scales spans the current invariant metric deformation space, and the exact Shape constraint system subsequently removes the physical tangent directions at the frozen point.

Quantum-design consumption#

The QC compiler treats the three root-plane families as a sparse interaction template. Rather than searching all pairwise couplings among local degrees of freedom, candidate checks/couplers are assigned to one of three symmetry-related interaction families. This reduces the topology search while retaining enough noncommutativity to mix sectors.

Load-bearing intervention#

Required cross-layer dependencies#

• K6 symmetry/Weyl ledger
• C_admiss candidate grammar
• Rigidity perturbation tests

11. Topology, cohomology and automorphisms - the part QEC already knows can be physical#

Physics function#

The Shape Stage carries exact topology and cohomology information, including the Poincare polynomial and Betti numbers of X9, plus a symmetry/automorphism ledger. In topological quantum error correction, homology classes are not descriptive decoration: they label logical operators and code sectors. This is therefore the least speculative cross-domain bridge in the dossier.

Worked physics example#

For the active internal Stage, b1(X9)=0 and pi1(X9)=0. That matters because it explicitly prevents the active quotient from silently supplying a topological one-cycle. The physics source keeps parent-cover cycles and active-quotient topology separate. This is exactly the kind of distinction a topological code compiler must preserve if logical sectors are to be trusted.

Quantum-design consumption#

The QC compiler converts Stage topology into a chain complex and computes candidate logical sectors from kernels/modulo images rather than from local stabilizer count alone. Automorphisms are used to quotient duplicate placements before expensive decoding simulation.

Load-bearing intervention#

Required cross-layer dependencies#

• Boundary incidence for relative homology
• Co-Actor global/topology failure class
• Granularity for finite chain-complex representation

12. Rigidity and the full-precision scalar packet - robustness instead of a razor-thin optimum#

Physics function#

Shape V8.1 distinguishes constitutive rigidity from dynamical stabilization. In the active metric/radius basis q=(log u1,log u2,log u3,log R2,log RY), five exact constraints have Jacobian rank five and determinant two, leaving zero physical metric tangent dimension under the declared Shape constraints. The unrestricted control still contains a negative mode m^2=-1/3; the source therefore does not pretend that a constrained direction is dynamically stabilized.

Worked physics example#

The worked physics point is methodological: the source explicitly carries both the full-rank constraint result and the unrestricted negative control. That is stronger than reporting only the favorable constrained calculation, because it tells a reviewer what kind of stability is actually being claimed.

Quantum-design consumption#

In the QC compiler, Rigidity supplies a perturbation test around each selected architecture. Let x parameterize couplings, check weights, phase settings and fabrication-sensitive control values; let C(x)=0 denote hard code constraints. The Jacobian and the smallest singular value of the relevant response map are used to reject solutions that sit on an artificially thin admissible needle.

Load-bearing intervention#

Required cross-layer dependencies#

• Scale for physical units and uncertainty
• Granularity for finite perturbation/stopping rules
• Interdependence for full downstream replay

13. F+ finite chamber - a finite operator geometry that actually generates structured outputs#

Physics function#

F+ is a finite, non-propagating Rulebook/operator chamber. Its current object includes the order-three modular point tau=omega, a three-dimensional generation module, sector projectors, four chamber operators, action ladders, phases, normalizations and a deterministic Yukawa map. It contributes zero metric dimensions but is explicitly load-bearing in the flavor claim.

Worked physics example#

The physics worked example is the quark pipeline. The up and down chamber operators are frozen from the action ladders; the two declared flavor calibrations are y_t(MZ) and |V_us|; the source then diagonalizes the generated Yukawa matrices and compares the resulting quark masses, CKM data, CP phase and Jarlskog invariant. Whatever one thinks of the underlying theory, the operator chamber is doing explicit mathematical work rather than serving as a label.

Quantum-design consumption#

The QC compiler uses F+ as a finite logical-sector chamber. The three-state generation module becomes a finite routing basis for three equivalence classes of syndrome/control responses; sector projectors enforce orthogonality; the order-three phase structure supplies a small discrete gate/phase alphabet; and family-level normalization knobs are forbidden for the same anti-fitting reason they are forbidden in the flavor construction.

Load-bearing intervention#

Required cross-layer dependencies#

• K6 family/Weyl structure motivates the three-sector basis
• C_admiss freezes the chamber before held-out comparison
• Matter/control Actors provide domains for the finite operators

14. C_admiss - the direct admit-one-eigenspace/reject-the-rest bridge#

Physics function#

C_admiss is the finite admissibility Rulebook and anti-fitting firewall. It owns what configurations are allowed, which objects are frozen before comparison, layer-smuggling prohibitions, and the downgrade/reopen discipline. The GUT and Quantum sources explicitly identify the shared engineering operation: admit one eigenspace of a constrained operator and reject the rest.

Worked physics example#

In the physics workflow, admissibility prevents a branch from being rescued by silently changing the metric, bundle, chamber, comparison ruler or parameter set after an observable has been read. It is a governance/Rulebook object, not a metric dimension.

Quantum-design consumption#

In QC-G1, C_admiss becomes the code-space and architecture firewall. The compiler first rejects candidates that violate commutation, domain, connectivity, error-correction or resource constraints; expensive decoder simulation is run only on candidates inside the admissible class. At the state level, the stabilizer projector is the exact eigenspace-admission operation.

Load-bearing intervention#

Required cross-layer dependencies#

• Co-Actors define rejected/failure classes
• Governance fixes the freeze-before-compare rule
• Granularity defines finite stopping/exhaustion

15. Sector projectors and invariant class connections - structure that survives renaming#

Physics function#

The finite chamber carries orthogonal sector projectors, while Shape V8.1 separately builds Actor classes, Co-Actor classes and a 13x18 Actor-Co-Actor envelope. The important idea is not the names of individual particles or devices but invariant relations: ownership, support, domain, complement, spectrum, interaction, rigidity and scale roles persist under a change of concrete identity.

Worked physics example#

In the particle-physics construction, the projectors separate up/down/lepton/neutrino sectors and the macro-projector identity is used in the no-mediator/FCNC story. The Actor ontology then says a field is not fully specified until support, owner, domain, quotient, spectrum, dynamics and other clauses are assigned.

Quantum-design consumption#

The QC compiler uses invariant classes to prevent implementation labels from becoming hidden parameters. Data-mode Actor A and data-mode Actor B must satisfy the same applicable clause set even if one is a cavity mode and the other a transmon. Co-Actor classes define the corresponding failure envelope. A valid design should therefore be invariant under relabelings that preserve the typed incidence graph.

Load-bearing intervention#

Required cross-layer dependencies#

• Actor registry
• Co-Actor registry
• Rulebook R-18 one-owner/equivalence rule
• Interdependence graph

16. Matter Actors - explicit ownership of the information-bearing degrees of freedom#

Physics function#

The current Shape Actor registry contains separate chiral matter modules for Q_L, u_R, d_R, L_L, e_R and nu_R, each with family multiplicity and an eighteen-clause witness envelope. The purpose is not merely particle naming: it prevents support, representation, domain, chirality, spectrum and interaction ownership from being smeared across unrelated terms.

Worked physics example#

The physics worked example is that the K6 family index and the orbifold chirality projector require an actual bundle/Actor space on which to act. Without E_matter, an index count and a parity operator are symbolic statements with no domain. This is why the source classifies the tensor/Actor layer as load-bearing rather than bookkeeping.

Quantum-design consumption#

The QC compiler maps this role to information-bearing data degrees of freedom. A data Actor must declare physical support, code-sector membership, measurement accessibility, leakage space, interaction ports and logical ownership. Syndrome ancillas and couplers are not allowed to masquerade as data resources, and vice versa.

Load-bearing intervention#

Required cross-layer dependencies#

• Stage support
• Rulebook carrier/domain admission
• Gauge/control Actors for interactions
• Co-Actors for leakage/error complements

17. Gauge Actors - interactions and syndrome/control channels must have an owner#

Physics function#

The gauge Actor layer contains separate SU(3)c, SU(2)L and U(1)Y connection modules. In the physics construction these are the force-mediating connections on the appropriate bundles; the geometry supplies carriers while the Actor layer supplies the connection/field content. This separation is central to the project rule that isometry is not identical to gauge symmetry.

Worked physics example#

The term dossier emphasizes that C8 is not self-sufficient: it consumes the geometry, matter reps, global quotient and threshold packet but does not itself derive family count, flavor or proton safety. That explicit dependency discipline is the transferable feature.

Quantum-design consumption#

In the QC compiler, gauge Actors become interaction/control/syndrome channels. Every multi-mode check must be realized by a named coupling channel with support, range, schedule and error model. The code graph can propose an abstract stabilizer, but the gauge/control Actor is what turns it into a physical circuit cost.

Load-bearing intervention#

Required cross-layer dependencies#

• M4 causal support
• Stage locality/incidence
• Co-Actor correlated-fault classes
• Dynamics for time evolution

18. Higgs/Wilson Actor - protected transformations via holonomy#

Physics function#

The electroweak/Higgs Actor is represented in the current GUT source as a Wilson-line/Hosotani mode with an integer winding label. The source claims one-loop protection for the named sector and explicitly downgrades higher-loop all-orders protection to diagnostic. That scope discipline is important: topological/holonomic protection can be real without implying unlimited immunity.

Worked physics example#

The physics worked example records n_H=1, a finite chamber determinant eta_BK, and frozen predictions for v and m_h under the declared assumptions, while making the loop-level claim boundary explicit. The key structural idea is that a transformation can be encoded in a global path/holonomy rather than a locally tuned scalar parameter.

Quantum-design consumption#

Holonomic quantum computation is an established external concept: adiabatic or otherwise controlled loops in a parameter manifold can implement logical unitaries through geometric holonomy on a protected subspace. The QC compiler therefore assigns the Higgs/Wilson Actor to protected logical transformations, separate from passive memory protection.

Load-bearing intervention#

Required cross-layer dependencies#

• Parent-cycle/control support
• C_admiss protected subspace
• Dynamics for adiabatic/nonadiabatic control
• Rigidity for loop-deformation robustness

19. Proton-safety projector - forbid catastrophic cross-sector channels by construction#

Physics function#

The proton-safety term encodes a no-mediator/sector-orthogonality structure. A representative identity in the source is Pi_q M Pi_l=0: the dangerous cross-sector block is forced to vanish at the operator level for the declared class. This is more useful for engineering translation than the specific baryon-physics interpretation.

Worked physics example#

The physics worked example is the distinction between suppressing a dangerous rate and eliminating an operator class. The source claims the latter only for a scoped class and treats numerical proton lifetime more cautiously. That operator-class scoping is exactly how the QC analogue should be written.

Quantum-design consumption#

In QC-G1, define P_L for the protected logical sector and P_E for a catastrophic leakage/correlated-error sector. The safety Actor imposes a design constraint that the first-order implementation Hamiltonian has no direct block connecting those sectors for the named interaction class. Errors can still occur through higher-order or unscoped channels; those remain Co-Actors.

Load-bearing intervention#

Required cross-layer dependencies#

• F+ sector projectors
• Interaction Actor
• Co-Actor correlated-fault and leakage classes
• Dynamics for higher-order pathways

20. Co-Actors - the negative space is part of Shape#

Physics function#

Shape V8.1 defines eighteen universal Co-Actor complement classes: parent validity, provenance, owner, type, domain, quotient, chirality, global topology/anomaly, measure, spectrum, observer, dynamics, interaction, rigidity, scale, canonical decomposition, regulator, and exhaustion. A Co-Actor is not merely "an error"; it is the concrete complement/falsifier of an Actor clause.

Worked physics example#

In the physics workflow, Co-Actors force a gate to name what would invalidate the Actor claim: wrong domain, missing global lift, mirror sector, anomaly phase, negative norm, hidden pole, causal violation, hidden scale, and so forth. This is the project-wide embodiment of "publish the failure mode."

Quantum-design consumption#

In QC design, Co-Actors are the explicit error/failure envelope. They include ordinary Pauli/displacement noise but also leakage, correlated faults, missing reset, nonlocal control assumptions, invalid boundary domains, decoder model mismatch, calibration drift, fabrication defects, and resource-accounting omissions. The compiler is not allowed to report a code without a typed complement set.

Load-bearing intervention#

Required cross-layer dependencies#

• Actor clause applicability
• Governance freeze of the error alphabet
• Interdependence for propagation of a newly discovered failure

21. The complete Stage + Rulebook + Actors object - why the layers must be tested together#

Physics function#

The current Shape ontology says the Stage alone is not the physical object. Stage supplies supports/topology/incidence; Rulebook supplies admissibility/domains/quotients/equivalence; Actors supply independently owned degrees of freedom and interactions; Co-Actors supply the complement/falsifier space. The GUT A2/B2 discussion similarly argues that a finite chamber without tensor domains is under-specified.

Worked physics example#

The physics worked example is chirality/family structure: K6 topology can supply an index, and the interval quotient can supply a parity projection, but those operations still require a matter bundle/domain on which to act and a Rulebook declaring the domain/parity. The claim is inherently cross-layer.

Quantum-design consumption#

QC-G1 is therefore compiled only from the complete typed object. A support graph without a check/admissibility Rulebook is just hardware connectivity; a check algebra without data/control Actors has no physical implementation; data/control Actors without an error complement have no hostile benchmark. The full Shape is a closure package rather than one geometry picture.

Load-bearing intervention#

Required cross-layer dependencies#

• All preceding components
• Governance for equal permissions
• Granularity for stopping criteria
• Dynamics for actual circuit evolution

22. Worked design gate: building QC-G1 from challenges, nature-derived tensors, and the frozen Shape#

22.1 Challenge ledger#

The proposed quantum architecture is not obtained by asking a black-box optimizer to search every code. It is constructed gate by gate. For each challenge the team records several natural mechanisms, extracts the invariant functional principle, expresses it as a tensor/constraint, routes it to a Shape component, and only then searches the reduced space.

22.2 Search-space reduction is useful, but it is not the proof#

The search-space reduction is an engineering benefit. It is not what makes Shape load-bearing. A component can be load-bearing even if a generic solver could eventually rediscover the same solution, provided that removing the component from the frozen Shape changes the compiled architecture or the cost of legal recovery. Conversely, a component that merely makes search faster but never affects the attainable solution class should be described as a computational accelerator, not as evidence that the physical geometry is unique.

22.3 Proposed architecture stack#

The current project corpus already describes a simulated hybrid stack and reports legacy memory-floor and decoder-channel metrics. Because the complete replay artifacts are not in the presently accessible bundle, this dossier treats those numbers as legacy simulated evidence requiring re-certification. The new compiler retains the same broad architecture pattern but makes resource ownership explicit:

23. The actual load-bearing proof protocol#

23.1 Intervention semantics#

For each component X, define a frozen baseline Shape S and an intervention I_X(S). The modified compiler is allowed to re-optimize every downstream degree of freedom that is not itself frozen by the experiment. It is not allowed to reintroduce X under a new name or to change the measured hardware/noise anchors.

Here M is the preregistered metric vector, not one scalar score. A component earns "load-bearing" only when Delta_X has the predicted direction on the named subsystem and the effect survives equal re-optimization.

23.2 Metric vector#

23.3 Placebo controls#

A valid ablation experiment must also include interventions that should not change physics or engineering performance: coordinate relabelings, Actor renamings, stabilizer-generator reordering, symmetry-equivalent Weyl permutations, and basis changes accompanied by the correct conjugation of operators. These should produce the same compiled architecture up to isomorphism and the same metric distribution.

23.4 Restoration controls#

Every destructive intervention is paired with a restoration run. The sequence is baseline -> ablate/substitute -> fully re-optimize -> restore -> recompile. A causal-looking signature is a reversible degradation in the predicted subsystem. A degradation that does not recover after restoration is more likely to be stochastic search drift or harness contamination.

23.5 Substitution controls#

Ablation tests whether information matters; substitution tests whether the specific information matters. K6 should therefore be compared with explicit rival carriers, the quotient with the unquotiented circle, the structured finite chamber with equal-parameter generic finite operators, and the root-plane interaction template with degree-matched random/alternative templates. The conclusion ladder is:

1. Redundant: removal changes nothing.
2. Useful: removal hurts but a simple substitute restores performance.
3. Load-bearing: removal causes a specific loss; nontrivial substitutes or added complexity are required to recover.
4. Cross-domain consilient: the same independently selected structure is load-bearing in physics and in QC under separate evidence chains.
5. Not established by this experiment: uniqueness of the geometry or theorem-level proof that nature must use it.

24. Upstream physics freeze: Standard Model and quark reconstruction#

24.1 Why the Standard-Model reconstruction belongs in this dossier#

The cross-domain argument is only interesting if the Shape was doing real work before the quantum example. The current GUT source therefore serves as the upstream record. This section does not independently certify the GUT; it summarizes the mechanism that was frozen before the engineering test so a reviewer can see what information was already present.

24.2 Gauge/family/chirality chain#

K6 carries the SU(3)/family stage, S2 supplies the weak carrier/T3 structure, and the orbifold interval supplies parity/fixed-boundary information. The matter and gauge Actors place fields and connections on those supports. The source keeps the global Z6 quotient, anomaly ledgers, and bundle data in separate authority rows rather than pretending the bare product manifold is sufficient.

24.3 Quark/flavor chain#

For the quark sector, the frozen chamber data include tau=omega, kappa=exp(-pi sqrt(3)), up/down action ladders, a sector-level normalization rule and a deterministic Yukawa map. The current quark appendix records y_t(MZ)=0.9665 and |V_us|=0.22436 as the two declared flavor calibrations for the overall construction, then treats the remaining masses/mixings as generated/comparison outputs under the stated assumptions.

24.4 Why this matters for the QC test#

The engineering compiler is not allowed to change tau, the K6 topology, the quotient action, the Actor classes, or the projectors in order to obtain a better QC benchmark. If it did, the experiment would no longer be a transfer test of the frozen Shape. The cross-domain value comes from asking whether the structure already carrying these physics responsibilities also constrains a useful code architecture.

25. Master component-to-quantum-design ledger#

This table is the compact statement the public-facing page can eventually use. The long chapters above are the technical authority behind each row.

25A. Parameter collapse: the computational consequence of a structured Shape#

Why parameter collapse is a separate load-bearing claim#

The component chapters above establish a route by which each piece of Shape can constrain a quantum design. A second, distinct claim is more computational: the geometry can collapse the number of independent coordinates that a topology or architecture search must explore. This is not the claim that Shape forces a unique quantum computer. It is the claim that the frozen structure identifies equivalences, forbidden directions, invariant sectors, typed interfaces and low-dimensional tensor families before an expensive search begins.

That distinction matters. A topology search can be impossible in practice even when the final design is simple, because the raw representation contains enormous redundancy. Two candidate graphs may differ only by relabeling; three coupling blocks may be copies related by symmetry; a dense operator parameterization may contain cross-sector matrix elements that the Rulebook later projects to zero; a continuous deformation may be gauge or coordinate rather than physical. Searching all of those coordinates treats the same physical design as many different candidates. Shape is useful if it removes those redundancies before optimization and does so without removing the true optimum.

Source status#

Claim ceiling#

Parameter collapse is beneficial only if the reduced parameterization remains sufficiently expressive. A smaller search space can be worse if it excludes high-quality designs. Therefore this Part defines two obligations that must be tested together: collapse (fewer independent candidates/coordinates) and coverage (the reduced space still contains strong solutions). A search-speed improvement without a coverage audit is not evidence that the Shape parameterization is correct.

25B. Formal search-space measure and effective dimension#

A mixed discrete/continuous definition#

Let a raw quantum-design description contain a discrete topology tau in a candidate set T and a continuous parameter vector theta in a p-dimensional region X_tau. Shape supplies exact constraints F=0, inequalities A>=0, typed-domain conditions D, and an equivalence relation generated by a symmetry or redundancy group G. The physically admissible search object is not the raw Cartesian product; it is the quotient of the admissible subset:

For a finite discrete search, the exact discrete collapse factor is simply the ratio of raw candidates to admissible equivalence classes. For continuous coordinates there is no finite count without a ruler, so the clean comparison uses a fixed external resolution epsilon. This deliberately prevents Granularity from receiving credit in the Shape-only test: epsilon is supplied identically to every representation.

Here N_epsilon is an epsilon-covering number under a preregistered metric. B_epsilon is conveniently read as the number of search bits removed by Shape at that resolution. The use of bits does not assert a fundamental information ontology; it is only an accounting device for comparing search sizes.

Local continuous dimension#

Near a regular point, if r independent equality constraints are active and the redundancy group has an orbit of dimension g_orb, the local physical search dimension is

This is the differential-geometric version of parameter collapse: equality constraints remove normal directions and the quotient removes tangent directions that change the coordinates but not the physical design. The regularity assumptions are load-bearing. Rank-changing strata must be handled separately rather than averaged into a flattering dimension count.

Why dimension matters computationally#

On a simple hypercubic grid with q samples per independent coordinate, exhaustive sampling scales as q^d. This is not a model of every optimizer, but it is an exact illustration of the curse of dimensionality. Collapsing from p coordinates to d effective coordinates changes the grid count by q^(p-d).

25C. Symmetry quotienting: Weyl orbits and tied quantum-design parameters#

Finite-group orbit collapse#

The safest parameter collapse is the removal of designs that are physically equivalent by an exact finite symmetry. If a finite group G acts on a discrete candidate set X, the number of inequivalent candidates is given exactly by Burnside's lemma:

When the action is free, no nonidentity group element fixes a candidate and the quotient reduces the count by exactly |G|. When fixed points exist, the reduction is smaller and Burnside's formula prevents us from simply dividing by the group order. This is the correct way to credit symmetry without double-counting highly symmetric architectures.

K6/Weyl worked example#

The frozen K6 factor has Weyl group S3 with six elements. Its three invariant root-plane metric weights (u1,u2,u3) are permuted by S3. A general invariant homogeneous metric may assign three positive weights, but the Weyl-invariant isotropic point satisfies

At the level of Shape ratios this collapses a three-coordinate positive cone to one common coordinate; when Scale is held external and fixed, the two independent squashing ratios are removed entirely from the Shape search. This is a direct parameter-collapse example already present in the frozen geometry: symmetry does not merely decorate the answer, it identifies directions that the selected branch refuses to tune independently.

Quantum-design translation#

The prospective QEC compiler can use the same rule to tie parameters of symmetry-related coupling/check motifs. If three root-plane families each carried q independently tuned real coefficients, an unconstrained parameterization would expose 3q reals. Enforcing the frozen S3 equivalence permits one q-dimensional template plus the discrete group action:

This is a factor-three continuous parameter collapse if the symmetry is truly respected by the hardware/noise model. If fabrication or noise explicitly distinguishes the three families, tying them would be an unjustified overconstraint. The load-bearing test must therefore include a symmetry-breaking noise arm.

25D. Representation-theoretic collapse: searching the commutant instead of all operators#

From representation theory to a smaller operator algebra#

Symmetry can collapse not only labels but entire operator spaces. Let a group representation decompose a Hilbert space as a direct sum of irreducible representation spaces V_lambda with multiplicity spaces M_lambda:

Any operator that commutes with the group action lies in the commutant. For ordinary complex irreducible representations, Schur's lemma gives the block form

A generic Hermitian operator on a d-dimensional complex Hilbert space has d^2 real parameters. An equivariant Hermitian operator has only the parameters in the multiplicity-space Hermitian blocks; the real dimension is sum m_lambda^2 under the standard complex case. This is a structural parameter reduction, not a numerical fit.

Exact S3 regular-representation micro-example#

The six-dimensional regular representation of S3 decomposes into the trivial irrep once, the sign irrep once, and the two-dimensional standard irrep twice. Thus the multiplicities are (1,1,2). A completely generic Hermitian 6x6 operator carries 36 real parameters; an S3-equivariant Hermitian operator carries

a six-fold parameter reduction in this exact toy representation. The numerical factor six is especially transparent because the same group order appears in the K6 Weyl structure, but the mechanism is representation-theoretic: the group forces repeated matrix elements and zeros in an appropriate basis.

Why this matters to topology search#

A QEC architecture search frequently alternates between discrete topology and continuous operator/control choices. If every candidate topology is followed by a dense 6x6 operator optimization, the inner loop is 36-dimensional. If the Stage/Rulebook certifies S3-equivariance, the inner loop can be parameterized directly by six real numbers. The outer topology search is unchanged, but every candidate becomes much cheaper to evaluate and less prone to finding physically equivalent variants.

25E. Projector and admissibility collapse: remove forbidden matrix directions before search#

Projectors turn a dense matrix problem into typed blocks#

The Rulebook and finite chamber contain sector projectors. In a basis adapted to mutually orthogonal projectors P_a with sector dimensions n_a, a generic Hermitian operator on total dimension n=sum n_a has n^2 real parameters. If the admissibility rule requires block preservation, P_a O P_b=0 for a!=b, the allowed operator has only

For three equal sectors of dimension m, the reduction is exact:

so two thirds of the dense matrix directions are removed before optimization. Those removed directions are precisely cross-sector couplings. If a safety projector forbids direct transitions between protected and catastrophic sectors, those zeros are simultaneously a physics/safety statement and a computational search reduction.

Admissibility as architecture pruning#

The same logic applies before any continuous optimizer runs. Let A(x) be a deterministic admissibility predicate for a candidate architecture x. Instead of allocating expensive simulation to every x in the generator output, the compiler evaluates A first and sends only A=1 candidates downstream:

If N0 candidates are generated and N1 survive, the exact discrete collapse is N0/N1 for that stage. No independence assumption is required. With multiple nested filters the survivor counts N0>=N1>=...>=Nk are directly measurable, and the total collapse is N0/Nk regardless of correlations between filters.

Connection to the Round-2 lesson#

The earlier computational tournament already suggested that explicit failure-envelope filtering can save expensive evaluations, but that result tested a particular executor rather than the fundamental geometry. The parameter-collapse experiment proposed here is stronger: each admissibility predicate must be traced to a frozen Shape component or a separately declared nature-derived tensor, and the expensive-evaluation savings are measured only after coverage and regret are checked.

25F. Topological parameter collapse: from all graphs to invariant-compatible complexes#

The raw graph space is enormous#

For n labeled hardware or syndrome nodes, a simple undirected graph has M=n(n-1)/2 possible edges, so the completely unconstrained topology space contains

Already at n=24 this is 2^276 labeled graphs. A topology-first quantum-design program therefore cannot rely on blind enumeration. Shape can help only if its invariants translate into necessary structural conditions that eliminate large classes without using the target benchmark score.

Exact Betti-number edge-count screen#

For a connected graph regarded as a one-dimensional cell complex, the first Betti number is beta1=m-n+1, where m is the number of edges. Requiring a target cycle rank b therefore fixes the edge count to m=n-1+b. This condition is necessary but not sufficient for connectedness and the desired topology, which makes the following count a conservative upper bound on survivors.

Take n=24 and b=2. Then M=276 and m=25. The raw labeled-graph space has 276 search bits. Restricting only to the required edge count leaves at most binom(276,25) candidates, whose base-two logarithm is approximately 117.415. Thus this single topological invariant removes at least 158.585 bits of labeled-graph search before connectedness, isomorphism, hardware locality, stabilizer commutation, distance, or fault-tolerance constraints are applied.

This calculation does not say that beta1=2 is the correct target for our final QEC architecture. It is an exact worked example showing why topological invariants can be computationally powerful: they convert a free edge search into a sharply restricted combinatorial shell.

Cell-complex generalization#

For higher-dimensional topological codes, incidence matrices boundary_k obey boundary_{k-1} boundary_k=0. Instead of searching arbitrary parity-check matrices, a compiler may search cell-complex generators whose chain-complex identity enforces commutation structurally. The search variables become cells, incidences, quotient/gluing data and a smaller set of weights rather than every matrix bit independently.

25G. Constraint compilation for CSS/QEC searches#

Raw CSS-matrix coordinates versus generated incidence coordinates#

A CSS stabilizer code can be represented by binary check matrices H_X and H_Z satisfying the symplectic commutation condition H_X H_Z^T=0 over F2. If one naively exposes every matrix entry as a binary design variable, the raw coordinate count is n(r_X+r_Z) bits before commutation, row-basis equivalence, hardware locality or distance are enforced.

A topological compiler can instead generate H_X and H_Z from incidence/boundary operators of a typed complex. Then commutation follows from boundary-squared-zero rather than being rediscovered by rejecting matrix pairs. This is a canonical example of constraint compilation: a theorem or structural identity is built into the parameterization so invalid coordinates never exist.

Why this is more than prefiltering#

Prefiltering samples an invalid point and throws it away. A compiled parameterization never proposes that point. This distinction matters when the invalid fraction is enormous. The ideal Shape-driven search therefore moves constraints as far upstream as possible: topology and group representation define the generator; Rulebook projectors define legal operator blocks; Actors define typed ownership; Co-Actors define negative controls; only then does numerical optimization choose remaining continuous coefficients.

Search accounting to record#

25H. Actor classes and Co-Actors: collapse by typed ownership and invariant relations#

Typing removes meaningless permutations of ownership#

The current Shape does not treat every object as an interchangeable node. Actors own different kinds of physical work: data/matter-like degrees of freedom, gauge/control connections, protected transformation channels, and safety projectors; Co-Actors own complements, leakage, quotient, rigidity and failure witnesses. In a design search this typing can collapse parameter permutations that differ only by assigning the same numerical module to physically inequivalent roles.

Let q resource instances be partitioned into k fixed role classes with counts q1,...,qk. If an untyped architecture generator redundantly explores every reassignment of those roles to labeled resources, the number of role labelings is the multinomial coefficient

A typed Stage/Actor compiler can choose the role partition once and search parameters within each role, rather than re-learning the same ownership relation for every candidate. This is parameter collapse only when the role assignment is independently justified; otherwise it is a modeling assumption that must be tested by allowing role-swaps in a control arm.

Invariant-class sharing#

If several Actors belong to the same invariant class for a particular rule, a class-level parameter may replace one independent parameter per identity. If N identities are partitioned into K invariant classes and a scalar coefficient is truly class-invariant, the coordinate count collapses from N to K. For q coefficients per identity the count changes from qN to qK.

The earlier Round-2 diagnostic found that the present binary Actor-Co-Actor applicability patterns are coarse and do not uniquely distinguish every Actor. That is not a reason to abandon class sharing; it is a reason to enrich the invariant relation with owner basis, carrier, representation/domain, witness status and constraint clause before using it as a parameter-sharing theorem.

Co-Actors as negative-space collapse#

Co-Actors collapse the search by representing ways a candidate can be invalid or unsafe before expensive evaluation. The crucial control is to keep a held-out error/failure suite. If Co-Actor pruning only removes candidates that fail the designer-visible error model but increases failure on hidden channels, the apparent collapse is target-loaded and must be rejected.

25I. Rigidity collapse: search only physical deformation directions#

Rigidity is the local differential audit of parameter collapse#

After discrete topology and symmetry reduction, a candidate may still have many continuous coordinates. The Rigidity block asks which infinitesimal variations are genuine physical deformations. Let C(theta)=0 be the active equality constraints and J=dC/dtheta their Jacobian. At a regular point, tangent deformations lie in ker J. Gauge/redundancy directions are then quotiented before physical deformation rank is counted.

If theta has p components, rank J=r and the independent redundancy image inside the tangent space has dimension g, the regular local physical dimension is p-r-g. A full-rank Jacobian can therefore certify that a set of apparent knobs is locally fixed, while a null vector identifies an unresolved design degree that should remain in the optimizer.

Search consequence#

A black-box optimizer given p raw coordinates wastes evaluations moving along constraint-normal directions that are immediately projected back and along gauge/equivalent directions that do not change the design. A rigidity-aware parameterization chooses a basis V for the physical nullspace and searches latent coordinates z:

This is especially useful for topology refinement: once a discrete architecture is selected, each local continuous optimization is carried out only in its physical deformation directions. If a rank change occurs, the candidate has crossed to a new stratum and must be reclassified rather than silently continuing with a stale latent basis.

Robustness rather than only speed#

The same Jacobian/Hessian information distinguishes a broad admissible basin from a razor-thin optimum. Parameter collapse should therefore be scored together with perturbation margin. A design that uses very few parameters because it sits at an unstable singular point is not an engineering success.

25J. Nature-derived tensors layered onto Shape#

Nature-derived tensors as a second stage of collapse#

The user-defined design method adds functional constraints discovered by comparing several natural systems that solve an analogous problem. The common mechanism is encoded as a tensor or invariant relation and attached to the frozen Shape. This is not evidence for fundamental geometry by itself; it is an engineering strategy for turning qualitative biological/physical lessons into auditable mathematical restrictions.

Suppose a raw n-by-n coupling matrix W is treated as n^2 independent real entries. A nature-derived mechanism may suggest that legal couplings lie in the span of r frozen basis tensors B_alpha determined by locality, compartment boundaries, redundancy or conservation:

The optimizer then searches r scalar coefficients rather than n^2 matrix entries. Shape determines where each B_alpha is allowed to have support and which Actors it connects; the nature-derived tensor determines the functional pattern within that support. The two sources of structure are therefore compositional rather than interchangeable.

Sequential challenge gates#

Let Omega_0 be the compiler space before engineering challenge gates. Challenge j adds a deterministic admissibility condition C_j. The spaces are nested:

For finite enumerated pools, no independence assumption is needed: if N_j=|Omega_j|, then the empirical collapse at gate j is N_{j-1}/N_j and the total is N_0/N_J. For generative spaces, the same accounting is estimated from frozen samples and accompanied by confidence intervals.

Examples of functional tensors to test#

25K. Hierarchical collapse: a compiler that shrinks the search before simulation#

The complete collapse ladder#

The strongest version of the design method does not ask one optimizer to discover topology, symmetry, role assignment, admissibility, safety and continuous control simultaneously. It compiles those decisions in an ordered ladder, while preserving a record of which upstream fact removed each candidate or coordinate.

Why the order matters#

Cheap exact constraints should act before expensive simulation. Topological identities, group orbits and projector zeros are exact and deterministic; they should not be rediscovered by a Monte Carlo decoder. Conversely, a noise-dependent performance preference should remain downstream and should not be promoted into a fundamental Shape rule. This separation prevents the search method from laundering benchmark information into the frozen geometry.

Log-space accounting#

Because discrete topology counts can span hundreds or thousands of bits, the ledger should report log2 candidate counts or covering numbers. For nested finite sets the collapse bits add exactly:

This additive ledger lets a reviewer see whether most of the reduction comes from a mathematically justified quotient or from a late, target-sensitive heuristic. The former is strong evidence of genuine parameter collapse; the latter is merely optimization.

25L. Worked six-sector parameter-collapse experiment#

Worked design study: structured six-sector logical-memory module#

This section defines a concrete, reproducible toy architecture search intended to test the parameter-collapse machinery before the full hardware campaign. It is not presented as the final quantum computer. The purpose is to make every reduction countable.

Raw parameterization#

Begin with six labeled sectors, motivated only as a convenient testbed for the six Weyl chambers. Let a Hermitian interaction/control matrix H act on one effective mode per sector. A generic H has 36 real parameters. Independently, allow a simple undirected inter-sector support graph with 15 possible edges, for 2^15 topologies. The raw mixed search is therefore a 36-dimensional continuous family over 32,768 labeled support graphs.

Shape collapse 1: S3-equivariant operator#

Impose the exact S3 regular-representation equivariance discussed above. The Hermitian operator dimension collapses from 36 to 6 real parameters.

Shape collapse 2: topology orbit quotient#

Let S3 act on the six sectors according to the frozen chamber action and quotient support graphs by this action. The exact number of orbits is computed by Burnside's lemma in the executable test rather than assumed to be 32768/6, because some graphs have nontrivial stabilizers.

Shape collapse 3: admissibility/projector constraints#

Declare protected and control sectors through frozen projectors and forbid direct protected-to-catastrophic blocks. Candidate supports violating the block-zero rule are never generated. Record the number of surviving graph orbits exactly.

Shape collapse 4: nature-derived sparse propagation tensor#

Add a preregistered propagation constraint limiting the number and arrangement of cross-module fault channels. This is explicitly an engineering tensor, not fundamental Shape. It is applied after the Shape quotient so its incremental effect can be separated.

What is measured#

The study is successful if the Shape parameterization reaches the same or a better valid Pareto frontier with materially fewer unique expensive evaluations, and if the gain disappears selectively when the relevant Shape constraint is removed. It fails if the reduced space has high regret, if the generic solver finds better valid architectures outside it, or if the alleged reduction consists mostly of relabeling already handled cheaply by the baseline.

25M. Production quantum-design search: proving collapse without assuming optimality#

From the toy certificate to the full architecture search#

The production experiment scales the same accounting to the actual modular QEC architecture. The raw generator is intentionally broad enough to include strong conventional motifs, while the Shape compiler generates a structured subfamily. Both arms receive the same physical noise, connectivity, resource and decoder models. The objective is not to make the Shape arm win; it is to determine whether the Shape subfamily is a high-quality low-dimensional chart on the useful part of design space.

Three search arms#

Equal-budget protocol#

Each arm receives identical wall-clock or simulator-call budgets, identical held-out seeds and identical access to standard mathematical preprocessing. The strong raw arm is permitted to discover an equivalent low-dimensional representation; if it does, that shows the collapse is a useful representation but not unique to the Shape vocabulary. The load-bearing claim concerns the information encoded by the Shape, not proprietary terminology.

Primary statistical outputs#

What would be a convincing result#

A persuasive result is not merely “Shape searched fewer points.” It is a multi-part certificate: (1) exact or reproducible counts show large collapse; (2) the compiled space retains the strong frontier with low regret; (3) ablation selectively expands the search or worsens efficiency; (4) restoration recovers the effect; (5) placebo transformations do not change it; and (6) independent replay obtains the same counts and rankings.

25N. Upstream precedent: parameter collapse in the Standard-Model/flavor construction#

The upstream physics already uses the same idea#

The proposed QEC compiler is not introducing parameter collapse from nowhere. The physics construction already relies on structured parameterization: the K6 invariant metric has symmetry-related directions; quotient/parity data remove sectors; projectors enforce representation domains; the finite chamber generates Yukawa structures from a restricted operator family rather than four arbitrary dense complex matrices. The cross-domain hypothesis is that the same kind of structural compression is useful in engineering.

Flavor example with an explicit claim boundary#

A general complex 3x3 matrix contains 18 real parameters. The diagonal chamber expression used as one stage of the current quark construction,

is specified by a normalization and the common hierarchy parameter at that stage, before the full mixing/operator machinery is applied. This is a dramatic local parameter collapse relative to an arbitrary matrix, but it must not be misreported as the complete flavor parameter economy. The current GUT audit explicitly retired the older “four inputs to twenty-two outputs” oversimplification and records additional injected reals and structural data. The honest lesson for QEC is methodological: structured geometry can replace arbitrary matrix entries with invariant coordinates, but every remaining coordinate must still be charged.

Why cross-domain recurrence would matter#

If independently frozen physics structure repeatedly produces low-dimensional, high-coverage coordinate charts in quantum engineering, that is stronger than observing one numerical coincidence. The same mechanism would be doing two conceptually different jobs: reducing arbitrary free parameterization in particle physics and reducing arbitrary architecture parameterization in QEC. That recurrence is the consilience target.

25O. Failure modes and hostile-review controls for parameter collapse#

How parameter collapse can fool us#

A constrained search almost always looks more efficient if one ignores what it excluded. The validation therefore treats overconstraint as a primary failure mode, not a secondary caveat.

The no-free-lunch interpretation#

Shape can improve search only by injecting prior information. The scientific question is whether that prior information was independently justified by the physics and whether it generalizes to held-out engineering conditions. If it does, the collapse is evidence that the Shape encodes reusable structure. If it does not, the smaller search space is merely a bias.

Null results that remain valuable#

• No collapse: a component may be physically meaningful but irrelevant to the chosen QEC design coordinate system.
• Collapse with high regret: the component overconstrains engineering even if it is useful in physics.
• Collapse reproduced by generic preprocessing: the information is useful but not distinctive to the Shape formalism.
• Collapse only under symmetric noise: the contribution is load-bearing only in that hardware/noise regime.
• Large collapse with low regret across held-out regimes: strongest evidence for a genuinely useful cross-domain coordinate system.

25P. Pre-registration and certificate format for the parameter-collapse claim#

Parameter-collapse certificate to publish for every production run#

Minimum positive result#

For a component X to receive a PARAMETER-COLLAPSE LOAD-BEARING label in the engineering dossier, all of the following should hold:

1. X removes a mathematically identifiable equivalence, forbidden direction, typed redundancy or constrained degree of freedom.
2. The reduction is measured as a discrete count, covering-number reduction, rank reduction, or exact latent-coordinate reduction.
3. The compiled space retains a preregistered performance frontier within tolerance of the strong raw baseline.
4. Removing X expands the search or raises evaluation cost in the predicted way after equal reoptimization.
5. Restoring X restores the collapse/efficiency effect.
6. Placebo transformations leave the result invariant.
7. The effect survives held-out hardware/noise perturbations within the claimed scope.
8. An independent replay reproduces the result from the frozen compiler and seeds.

Stronger cross-domain label#

A component may be called CROSS-DOMAIN LOAD-BEARING only if its upstream physics role is independently certified at the claimed level and its downstream QEC parameter-collapse or performance role passes the engineering protocol. This wording still does not mean “unique geometry proven.” It means the same frozen structural information carries demonstrable load in two independent chains.

25Q. Narrative synthesis: Shape as a coordinate system for difficult design searches#

The scientific excitement is in the scoreboard that follows. If the reduced formation reaches the same or better Pareto frontier with orders-of-magnitude fewer unique expensive trials, and if removing the specific Shape structures gives those efficiencies back, the geometry is doing measurable engineering work. If the unconstrained team simply finds better plays outside the Shape formation, the geometry has overconstrained the problem and the result belongs in the loss column.

That is why parameter collapse is such a useful cross-domain test. It does not require the geometry to force a unique solution. It asks a more modest and more falsifiable question: did the geometry correctly identify coordinates that never needed to be searched independently? Every exact yes is a piece of load-bearing evidence. Every no tells us which claimed structural relation does not transfer.

Public-facing statement if the experiment succeeds#

26. Hostile-review objections and what would actually answer them#

26.1 "You designed the QC analogy after seeing the answer"#

Answer: freeze the component-to-compiler mapping, objective, hardware/noise model and held-out tests before running the final simulation. Preserve hashes. Any later improvement is a new branch and cannot be credited to the frozen test.

26.2 "These are generic mathematical ideas, not evidence for your geometry"#

Partly correct. Projectors, topology, holonomy and symmetry are widely used in QEC. The experiment becomes informative only if the specific combination and detailed frozen structure selected in the upstream physics provides predictive engineering constraints, and component-specific ablations produce component-specific losses. Generic usefulness is not enough.

26.3 "A sufficiently powerful optimizer can rediscover all of this"#

That does not automatically defeat load-bearingness. If the optimizer is given the same structural information under another representation, then a tie shows representational equivalence. If the information is genuinely removed and the optimizer reconstructs an equivalent structure from independent data, then the component is substitutable. The correct conclusion is not unique forcing. The experiment is designed to distinguish those cases.

26.4 "You can make any component look useful by choosing a favorable metric"#

Use a vector of preregistered metrics and publish all of them. A component earns load-bearing status only on the subsystem effect predicted before the run and only if the result survives the common-ruler resource ledger. No single weighted aggregate may hide a regression elsewhere.

26.5 "The physics claims themselves are disputed"#

The QC experiment does not cure that. It is an independent cross-witness. If a GUT gate fails later, the engineering result remains an engineering result; if the QC design fails, the physics certificate cannot cite it as a rescue. The no-feedback firewall is binding in both directions.

26.6 "A simulation is not hardware"#

Correct. Simulation can establish mathematical consistency, code properties under declared noise, and comparative resource projections. Hardware validation is a later evidence tier. The report therefore uses SIMULATED, REPLAYED, and HARDWARE-VALIDATED as distinct labels.

27. Evidence and claim ladder#

27.1 Evidence tiers#

27.2 Decision rule for each Shape row#

Each component ends in one of five states: NOT-USED, USED-NOT-LOAD-BEARING, SEARCH-ACCELERATOR, LOAD-BEARING-SUBSTITUTABLE, or LOAD-BEARING-DIFFICULT-TO-SUBSTITUTE. The public narrative should print the state next to the component rather than converting every result into a success.

27.3 What would be genuinely surprising#

The most compelling outcome would not be a single headline ratio. It would be a pattern of orthogonal, preregistered effects: K6 symmetry changes routing/redundancy; the quotient changes leakage; Co-Actors change correlated-fault resilience; the safety projector changes catastrophic propagation; Rigidity changes fabrication sensitivity; and placebo relabelings change nothing. A multi-component pattern is harder to explain as one lucky code optimization.

Appendix A - Full-precision Shape packet and unresolved convention#

A.1 Stage constants carried by the current Shape source#

A.2 Flavor/chamber constants used in the upstream physics#

A.3 Scale conflict that must not be hidden#

The Stage source gives Vol(IY)=pi RY. With RY=R0/2, this is 2.5e-17 GeV^-1 and Vol(X9)=1.852208630699352e-148 GeV^-9. A legacy SG-1/GUT numerical table prints twice those values, corresponding to RY=R0. This is an exact factor-of-two convention conflict. Until ratified, any engineering translation that needs an absolute interval scale must show both branches or remain dimensionless.

Appendix B - Current Actor and Co-Actor registries#

B.1 Primary Actor registry#

B.2 Universal Co-Actor class registry#

The full Actor-Co-Actor matrix contains 13 x 18 = 234 cells. This dossier uses the matrix as an applicability/provenance aid but does not treat a binary applicability pattern by itself as sufficient evidence of invariant structure. Rich witness fields and named systems are required in the final QC replay.

Appendix C - Exact micro-certificates used by the compiler#

C.1 Stabilizer/code-space projector#

Verified numerically during dossier assembly: idempotence residual 0.0; trace 2.0.

C.2 Parity projector#

Verified numerically: idempotence residual 0.0; trace 2.0.

C.3 S3 symmetrizer#

On three qubits, the exact permutation symmetrizer has trace 4; the numerical idempotence residual used here is 0.0. This demonstrates how a Weyl/permutation equivalence can be turned into an actual code-space restriction rather than a labeling convention.

C.4 No-direct-path block#

For a block-diagonal toy Hamiltonian the cross-block norm is exactly zero. Adding an explicit off-block epsilon supplies a continuous falsification knob. This is the miniature version of the proton-safety-to-fault-firewall translation.

Appendix D - Source register and reproducibility requirements#

D.1 Internal source register#

• shape(5).zip - Shape V8.1 universal Stage/Rulebook/Actor/Co-Actor package and registries.
• shape.html - integrated full Shape authority assembled in this project.
• GUT.md - current full-precision 13D instance, term dossiers, Standard-Model and flavor/quark source authority used here.
• QUANTUM_review_bundle.zip / pages/Quantum-paper.html - current engineering-bridge claim boundary and legacy simulated QC record.
• rigidity(5).zip, interdependence(3).zip, governance(3).zip, granularity(3).zip, scale(5).zip - supporting building blocks used only where explicitly identified.

D.2 Standard external QEC references used for context#

• D. Gottesman, Stabilizer Codes and Quantum Error Correction (1997 thesis / arXiv quant-ph/9705052).
• E. Dennis, A. Kitaev, A. Landahl, J. Preskill, Topological quantum memory, Journal of Mathematical Physics 43 (2002).
• D. Gottesman, A. Kitaev, J. Preskill, Encoding a qubit in an oscillator, Physical Review A 64, 012310 (2001).
• P. Zanardi, M. Rasetti, Holonomic quantum computation, Physics Letters A 264, 94-99 (1999).
• These references establish that stabilizer projectors, topological logical sectors, oscillator codes and holonomic gates are standard QEC/QI mechanisms. They are cited so the cross-domain argument does not misrepresent those ideas as inventions of the present Shape framework.

D.3 Reproducibility rule#

A final public version should ship with: the exact Shape source hashes above; the compiler source; generated candidate registries; the complete hardware/noise manifest; frozen seeds; full code/circuit/decoder outputs; ablation/restoration runs; placebo runs; and a machine-readable component-to-output dependency graph. Any result that cannot be regenerated from that package should be downgraded to narrative-only.

Appendix E - Complete Shape registry carried into the cross-domain test#

E.1 Complete current Stage-instance registry#

The table below is source-derived from Shape V8.1. It is included so a reviewer can verify that the QC compiler has not quietly reduced "Shape" to the four-factor product label. Every row is a potential dependency or ablation surface.

E.2 Complete current Rulebook-instance registry#

Likewise, the Rulebook is broader than the finite flavor chamber. The current Shape source classifies domain, quotient, anomaly, positivity, spectrum, dynamics, rigidity, ruler, observer, interaction, equivalence, regulator and evidence/freeze rules as part of the complete typed object.

E.3 How these registries are used in the QC replay#

A QC run does not need to instantiate every physics-specific rule literally. It must, however, disposition every row: CONSUMED, TRANSLATED, NOT-APPLICABLE-BY-GRAMMAR, or DEFERRED-TO-DOWNSTREAM-GATE. A row may not disappear silently. This is the same no-smuggling discipline the Shape source applies to physics.

Appendix F - Pre-registered component-by-component ablation, repair, placebo and null rules#

28. Conclusion - what would count as a convincing result#

What this dossier establishes today#

• The current Shape can be written as a complete typed object whose individual components have mathematically explicit physics roles.
• The existing GUT/Quantum corpus already contains a narrow, fenced QC bridge for K6 plus the admissibility chamber operation.
• A coherent prospective compiler can consume every major Stage/Rulebook/Actor/Co-Actor component without changing the upstream physics Shape.
• Several translations have exact micro-certificates and established QEC analogues: stabilizer projection, parity-sector projection, topology/homology, symmetry projection, block-off-diagonal safety, and holonomic control.
• The parameter-collapse Part now supplies a formal search-space framework: orbit quotients, projector/block reductions, topology-count reductions, class-level parameter sharing, rigidity/null-space coordinates, and nature-derived tensor parameterizations are all expressed as countable or rank-testable reductions rather than as a qualitative efficiency claim.
• The component-specific ablation/restoration tests are now defined in a way that can genuinely demonstrate load-bearingness rather than merely showing that labels are useful.

What remains to earn the strong public claim#

1. Freeze the Shape-to-QEC compiler and component ledger before new benchmark results are loaded.
2. Implement QC-G1 end to end with complete physical resource accounting.
3. Publish the parameter-collapse certificate: raw design grammar, quotient/orbit counts, effective continuous dimension, compiled latent variables, coverage/regret, and simulator-call savings under the strong raw baseline.
4. Run every component ablation with equal re-optimization permissions and held-out seeds.
5. Run restoration and placebo controls.
6. Run strong substitutions for the components where a natural rival exists.
7. Have an independent group replay the package.
8. Only then publish rows as load-bearing, substitutable, redundant, or unresolved.