# Full Text: Observers Who Read Their Universe into Existence: Finite Consensus, an Executable Observer, and Standard Physics

> Extracted from `ObserversReadUniverse_r2052.pdf`

---

## Page 1

Observers Who Read Their Universe into Existence:
Finite Consensus, an Executable Observer, and Standard Physics
Bernhard Mueller 1,2,* Mario Poneder Alexander Osika 3 Kai Xue
Ben Cassie Peter Nguyen Jinwook Kim 4 David Matscheko
Daniel Ari Friedman 5,6 Jonathan Hill William T. Glynn Maarten Antonie Visser
Kale Arnav Anirudha Brieuc de La Fournière
October 21, 2026 · Paper release: r2052
1Pragma Research Inc. 2Information Physics Institute 3EtherWorks 4Oraclizer Labs 5Active Inference
Institute 6FractiAI
*Corresponding author: bernhard@floatingpragma.ai
Abstract
Observer Patch Holography (OPH) proposes that physics is what bounded observers can
establish together: ﬁnite patches compare records on shared boundaries, repair disagreements,
and treat as objective whatever survives every comparison. The universe, on this view, is a
timeless self-referential structure whose observers recover its speciﬁcation from their records
and realize it, so that it reads itself into existence. Nothing lies outside to impose premises,
so the recovered speciﬁcation owes no a-priori justiﬁcation; it has to run and reproduce what
observers measure.
That speciﬁcation is a candidate, an executable observer: a twelve-port carrier, a native
repair law, retained records and record-controlled feedback, joined in one interacting quantum
process of electric ﬂux, charged matter and a scalar ﬁeld whose feedback response is certiﬁed
nonzero in the inﬁnite-occupation limit. Observers whose regions overlap on one carrier share a
single complete law of ﬁelds, local event clocks and feedback, and on a reﬁning spherical screen
one repair history deﬁnes a common continuum operator for Gaussian ﬁelds, their decay and
ﬁnite-speed waves. On a declared radial product over that screen, every ﬁnite menu of observer
readouts admits, to any tolerance, one smooth static (1 + 3) -dimensional Lorentzian eﬀective
ﬁeld theory. Finite theorems turn the speciﬁcation’s readouts into familiar structure: schedule-
independent public records, the Born rule and Tsirelson bound on completed records, a carrier
whose slow response is exactly three-dimensional, the Standard Model gauge Lie algebra with an
anomaly-free generation, a source-derived (1 + 3) causal precursor with proper-time ratios from
record counts, and the Einstein equation from a realized null balance. We state the acceptance
criteria of the construction, which premises it discharges, which it takes as declared inputs, and
the frozen tests that can reject its branches.
Interpretation and state variables. OPH proposes one structure with two descriptions: a
timeless, self-referential complete history and its actual internal computation. Observers are pro-
posed to reverse-engineer their reality’s substrate and construct hardware whose execution realizes
that same history, including its builders and hardware, in emergent, mutually calibrated local times.
The candidate must supply construction, initialization, controls, memory, read/write and record
services, clock comparisons, and their resource costs; no external host or additional time is assumed.
Its realization map must specify whether it matches one actual history, a family of preparation and
1

## Page 2

intervention responses, or every independently allowed microstate. These are diﬀerent requirements.
A candidate must specify the history domain and closure operator; neither the schematic equation
nor the ﬁnite constructions establish this physical realization or select a unique universe. An actual
quantum record history also need not assign simultaneous outcomes to incompatible unperformed
measurements.
The proposed primitives are patch states, records, interfaces, and admissible operations. Physi-
cal space, duration, and matter are intended emergent readouts. A supplied carrier incidence, geo-
metric domain, action, representation, or clock is a premise of the construction using it. Complete-
history closure, an instantaneous repair ﬁxed point, a public normal form, and a stationary prob-
ability law are diﬀerent objects. A stationary law can support changing individual histories. An
operational observer is deﬁned by its access and record interfaces; consciousness requires an addi-
tional psychophysical interpretation.
Time conventions. Host runtime, solver counters, native attempt counts, local event-clock read-
ings, event precedence or logical numbering, model coordinates or lapse, modular parameters, and
calibrated observer-clock readings have distinct roles. Their identiﬁcation requires explicit maps.
A modular parameter is dimensionless; its geometric and operational readings require the stated
branch and clock premises. Consistent local clock comparisons require no preferred universal clock
or identical clock rates. Shared time parameters in restricted benchmarks are declared model
choices. Experienced succession and a ﬁnal instantaneous cosmic endpoint do not follow from a
modular-ﬂow or ﬁnite-convergence theorem.
Keywords: foundations of physics; Observer Patch Holography; observer consensus; executable
observer; self-reference; quantum records; gauge structure; Einstein ﬁeld equation.
1 Introduction
Physics usually begins with a world and asks how observers ﬁt into it. Observer Patch Holography
(OPH) begins with observers and asks what world they can establish together. Its observers are
ﬁnite. Each holds a patch of local state, reads a boundary it shares with its neighbors, keeps durable
records, and repairs its state when records disagree; it accepts only repairs that preserve protected
data. OPH identiﬁes an objective fact with a record that survives every authorized comparison.
Quantum probability, Lorentzian kinematics, gravitational dynamics and gauge structure enter
standard physics through distinct mathematical inputs [ 17, 18, 19, 53, 60]; OPH asks whether
they are diﬀerent public readouts of one ﬁnite observer architecture. Space, time, matter and
classical objects are eﬀective descriptions of what observers can measure and jointly interpret; no
observer-independent continuum is assumed beneath them.
The proposal about the whole is closure. The universe is a timeless, self-referential structure
whose observers recover the speciﬁcation of their own world from completed records and build
hardware whose execution realizes that same history, including its builders. The structure reads
itself into existence. This changes what a premise has to do. A theory set against an external
world must justify its postulates in advance or by ﬁt; a structure with no outside has no place from
which a postulate could be imposed. Whatever observers recover and implement therefore needs no
further justiﬁcation: the carrier, the observer algebra, the repair law and two closure coordinates
are justiﬁed by being recovered, by running, and by reproducing what observers measure. The
same rule cuts the other way. Nothing may be assumed that observers have not recovered, and
every change of description they can make must be implemented by the structure’s own operations.
Agreement ﬁxes what counts as shared, so “our universe” is the world described by the public
2

## Page 3

record. The executable speciﬁcation is where this account enters the loop; the loop itself has no
ﬁrst step. Section 6 develops the hypothesis and its consequences.
The programme’s target follows from this reading. An executable construction should contain
observers who make the measurements we make and who can use the eﬀective continuum descrip-
tions we use: quantum ﬁeld theory, classical mechanics and dynamical gravity, each in its regime.
The speciﬁcation exists and runs. Its ﬁnite observer executes repair, writes and retains records
and chooses its next interaction from those records; an electromagnetic detector evaluates the cou-
pling from a closure coordinate and feeds a single bit back; and one interacting quantum process
joins electric ﬂux, charged matter and a scalar ﬁeld with complete measurements and a feedback
response certiﬁed nonzero in the inﬁnite-occupation limit. Two observers whose regions overlap
on one carrier read local ﬁelds and event clocks from the same continuing history, and a direct
native execution and a derived eﬀective description assign their joint records the same complete
law. Joined carriers form spherical screens, and on a reﬁning family of screens one native repair
history deﬁnes a continuum spatial operator that governs Gaussian density ﬁelds, their decay and
admitted waves with a ﬁnite propagation speed. Observers of that history can use a Lorentzian ﬁeld
theory as their eﬀective description: on a declared radial product over the screen, one smooth static
(1 + 3)-dimensional metric reproduces any ﬁnite menu of their joint clock, ﬁeld and response read-
outs to any prescribed tolerance. Section 4 turns seven readouts of the speciﬁcation into familiar
structure: schedule-independent public records, the quantum identities of completed records, three
spatial directions, the gauge algebra with one anomaly-free generation, a continuum ﬁeld operator
on the screen with its eﬀective (1 + 3) Lorentzian description, a causal precursor with record-count
clocks, and the Einstein equation up to a cosmological constant.
The results have diﬀerent logical types, and Table 1 gives the classes used throughout; each
result below carries one of them. Each step is identiﬁed by its artifact as a Lean theorem, exact
executable certiﬁcate, conditional paper argument, numerical receipt, or explicit unproved premise.
Throughout, “source” means computed from the declared ﬁnite source speciﬁcation, including
every explicitly named source law, with no measured physical input. A physical realization map is
not source-only merely because the ﬁnite branch calculation consumes no measured target value.
Deﬁnition 1.1 (Physical realization map) . A physical realization map assigns ﬁnite public records,
event data, currents, modular generators, and scales to physical records, spacetime events, labo-
ratory currents, clocks, and units, preserving the algebraic operations, overlap restrictions, causal
order, normalizations, and reﬁnement relations consumed by the conclusion in question. A resem-
blance of spectra, dimensions, or symmetry names is insuﬃcient.
How this paper is organized. This paper gives the account and develops selected results in
detail. Section 2 describes the observer, its carrier and repair law, and what runs. Section 3
shows what observers agree on: public normal forms and the quantum identities of completed
records. Section 4 develops six results in detail: the three spatial directions, the gauge algebra
and one generation, measurement inside one interacting process, the continuum operator of one
repair history and its eﬀective Lorentzian description, the source-derived causal precursor, and the
Einstein equation; it lists the other results brieﬂy. Section 5 states the completion target, the
acceptance criteria and their status. Section 6 develops closure. Section 7 gives the frozen tests,
Section 8 the veriﬁcation stack, and Section 9 the comparison with neighboring approaches. A
companion construction paper states the executable speciﬁcation and its certiﬁed components in
full [ 66], and the theorem library carries the long proofs [ 1, 4, 5, 6, 8, 11, 9, 7, 2].
Boundary 1.2. This paper claims no completed derivation of particle masses, no source-only de-
termination of the ﬁne-structure constant, no completed cosmology, and no claim that formal
3

## Page 4

Table 1: Epistemic claim classes used in the manuscript.
Finite theorem A deductive result proved from displayed premises on a stated ﬁnite
domain. An exhaustive computation counts as part of the proof only
with a speciﬁed domain and a soundness argument.
Certiﬁed ﬁnite
computation
An executable witness whose inputs, controls, and output are ﬁxed
by a reproducible certiﬁcate; execution alone does not promote its
interpretation to a theorem.
Conditional analytic
theorem
A theorem whose hypotheses include a supplied continuum action,
state space or limit; its conclusion holds on that supplied object.
Source limit theorem A theorem about the limit of a declared family of ﬁnite source
executions along a stated subsequence, with its normalizations and
order of limits displayed; the limit is derived, and its physical
identiﬁcation is a separate claim.
Reconstruction
implication
An exact implication whose antecedents include displayed geometric
or physical premises.
Unsupplied realization
map
A required map from a ﬁnite object to a physical event, current,
ﬁeld, clock, scale, or continuum structure for which no construction
is asserted.
Diagnostic A certiﬁed computation compared against measured values that it
consumes; never counted as a prediction.
Prospectively ﬁxed
conditional test
A diagnostic with a custody-bound target deﬁnition, kill band,
precision ﬂoor, and decision rule ﬁxed before comparison.
F rozen prospective
branch prediction
A source calculation and decision rule ﬁxed before any eligible
comparison, conditional on a named physical branch. Failure rejects
that branch; its scope extends to the whole framework only if the
branch is proved forced and exclusive.
4

## Page 5

veriﬁcation establishes physical truth. Quantitative comparisons that consume measured values
are labeled diagnostics. Section 7 states the frozen tests and the physical premises that set the
scope of any verdict.
2 The observer
2.1 Patches, records and repair
Deﬁnition 2.1 (Observer patch). An observer patch is a ﬁnite object with local state, an observable
boundary map, durable records, readback, and a set of authorized repair moves. A patch sees a
fragment of the world: its own state and the boundary data of its authorized overlaps.
Deﬁnition 2.2 (Overlap and repair) . For patches xi, xj with an authorized overlap e = (i, j ), both
induced boundary records must agree for the pair to hold public data. A repair move changes local
state while preserving protected readout. A conﬁguration is a normal form when no authorized
repair applies.
An observer in this sense is deﬁned by its access and record interfaces. It reads, writes and
retains records, and it uses retained records to choose its next operation. Nothing in the deﬁnition
refers to a background space, a global clock or a pre-assigned list of particles. Those enter later as
readouts that observers can share.
2.2 The twelve-port carrier
Each local carrier has twelve primitive boundary ports with oriented icosahedral incidence: thirty
seams and twenty faces, ﬁve seams at each port, every seam on two oppositely oriented faces.
Carriers federate through typed seam algebras with coherent triple overlaps into a nerve carrying a
degree-one bridge to an oriented spherical support. The theorem library calls this carrier, federation
and response layer A1. Within the closure hypothesis of Section 6 it is the part of the speciﬁcation
that observers recover from their records; every result below displays its dependence on it. The local
carrier Gram is recovered from native current records, while cross-carrier routing in the executed
federations uses a supplied cell-centre and tangent-frame convention.
Two exact features of the carrier enter later constructions. First, the proper rotation group of
the icosahedral packet is the alternating group A5, with sixty rotations acting on the twelve ports,
and the real port-coeﬃcient space decomposes without multiplicity as
P12 ∼=A5 1 ⊕ 3 ⊕ 3′ ⊕ 5. (1)
On the six-dimensional golden sector the spectral projectors have entries in 1
20 Z[φ], φ = (1 +√
5)/ 2, each is the entrywise Galois conjugate of the other, and both triplets are irreducible real
representations of real type; the projectors, characters and irreducibility certiﬁcates are machine-
checked [48, 1]. Second, record keeping is integer valued: a write appends +1, a retraction appends
−1, and readback sums atomic events per port. A conservative repair transfers one unit across a
seam whenever the oriented mismatch has magnitude at least two. With V (x) = ∑
i x2
i , each such
repair strictly decreases V by 2(d − 1) for mismatch d ≥ 2, so repair terminates. The native repair
law of Section 2.3 also moves load to the ﬂoor and ceiling of the endpoint mean, but at diﬀerence
one it waits or swaps by its coin, so the native process keeps moving while its quadratic cost stays
ﬁxed.
5

## Page 6

2.3 The native repair law
The executable observer runs on one carrier with twelve nonnegative integer loads of total two.
There are 78 such conﬁgurations, and the 66 binary conﬁgurations form an invariant shell. For an
oriented seam (a, b ) and a fair placement coin c ∈ {0, 1}, the native repair law moves
d = xa − xb, j = 1
2
[
d − (2c − 1)(|d| mod 2)
]
, y = x + j (eb − ea), (2)
so load is conserved and the signed current j is the native observation. Conservation, one-step
balance, a tie-break that reads no load and endpoint symmetry force this rule. Its quantum version
is a record law, and seam locality, the classical law and local number-phase covariance leave a convex
family of them, each a table of ﬁve coherence multipliers [ 68]. The executed law C records coin and
current, with Kraus operators L(c,j ) = 2 −1/ 2 ∑
x: j(x,c )=j |y(x, c )⟩⟨x|. Dropping the coin record gives
the most coherent member; adding the seam parity gives E, and E and the seam-total recorder are
the only idempotent members. Under a common deﬁnite-current record interface all members share
every classical current statistic. One coherent interference experiment reads the multiplier g that
separates C, E and the most coherent member (Section 7), and nine source-executed measurements
read every real coordinate of the table (Section 2.4). The quantum relaxation of the channels
obtained by discarding the source record, the forgotten-source channels, is known exactly. C and
E have quantum spectral gap (5 −
√
5)/ 60 on the binary shell and on the full carrier, their slowest
binary coherences decay by the sharp factors 5/ 6 per attempt for C and 7/ 10 for E, the binary-
shell channel becomes entanglement breaking after ﬁnitely many attempts ( 134 attempts suﬃce
by an explicit certiﬁcate), and on the full carrier it never does: an explicit partial transpose stays
negative at every attempt count [ 67, 69]. The endpoint instruments also ﬁx a local action. On
binary operator space Hnative = ∑
e(I −Φ e), summed over seams with Φ e the unread C or E channel,
is positive and local, with an exclusion exchange term and coherence-killing coeﬃcients set by the
record law; the uniform discrete transfer is I − Hnative/M for M seams. It is a Liouville operator
on operators, not a Hamiltonian on carrier states, and the executed channel strictly attenuates a
coherent mode, which rules out unitary conjugation on this carrier. At ﬁxed selected-pair sum each
classical repair minimizes the pair’s quadratic load cost, so ∥x∥2/ 2 falls by (d2 − (|d| mod 2)) / 4 per
repair [ 81, 67].
The gap and the arrow do not depend on the record law. Every admissible coherence table has
the uniform binary state as its unique stationary state and the gap (5 −
√
5)/ 60; the linear load
statistics along the rank-three port vectors of Theorem 4.1 attain it, while every coherence between
distinct load conﬁgurations contracts by at most 43/ 60 + 2
√
2/ 15 < 1 − (5 −
√
5)/ 60 per attempt,
so the longest-lived datum of a carrier is a classical position-like readout and equilibrium selects no
record-law coordinate [ 68]. On the binary shell each native repair is the ﬁbre-resampling map of
the theorem library for the uniform reference: it resamples the seam orientation uniformly within
the ﬁbre of its repaired visible datum, which minimizes relative entropy over that ﬁbre, while
a collision is resolved by deterministic descent. The library’s second law and modular Clausius
inequality therefore hold for every native attempt on the binary shell [ 8]. One such projection is
idempotent; the declared uniform schedule over noncommuting seams supplies a sustained arrow
and contracts every input to the uniform state at the gap rate, and passive repair never realizes
the library’s third map, the modular ﬂow of a faithful reference.
2.4 What runs
The speciﬁcation is executable, and its components run with retained receipts [ 67]. The companion
construction paper states each component, its declared services and its certiﬁcates [ 66]; the following
summary gives the parts used in this paper.
6

## Page 7

An observer that remembers and acts. The native observer executes two repair events with
a ﬁnite tape. It records the ﬁrst current, issues a forecast from that record, chooses the next seam
from the record alone, records the result and halts. Exhaustive checks retain all 264 binary-source
paths, and a 20, 000-trial seeded execution reproduces the exact predictions. Record-controlled
feedback gives ﬁnal port-zero occupation 35/ 264, against 1/ 6 when the selector is disconnected or
shuﬄed with the same action frequencies, so the response is caused by the retained record and
not by the action statistics. Feedback preserves the committed classical record while it can reduce
coherence between diﬀerent record values. Complete records of these short histories recover the
source optimally, with entanglement ﬁdelity D−2 ∑
h wh rank(Ph)2 for the maximally mixed input,
where D is the number of carrier conﬁgurations, h runs over the complete admitted histories,
wh is the product of their seam-selection and coin probabilities and Ph projects onto the inputs
compatible with h; it equals 1301/ 2028 for the full-carrier two-event C observer and 3329/ 6084 for
E under the same current-sign selector [ 70].
An electromagnetic detector driven by the closure coordinate. A second observer evalu-
ates the coupling α(P ) from the screen grain P and sets g(P ) =
√
4πα(P ). A prepared path qubit
reads the ﬁrst current through a calibrated loop exposure χ; the detector bit y = ±1 has prob-
ability [1 + y sin(g(P )j1χ)]/ 2 and alone chooses the second seam, giving mean transferred charge
E[g(P )j2] = (5 g(P )/ 264) sin(g(P )χ). All 528 weighted histories are retained and a 200, 000-trial ex-
ecution agrees with the analytic response. The closure coordinate thereby enters an actual response
magnitude; the charge assignment, probe, exposure and controller are declared apparatus.
Fields measured by the observer. A scalar detector prepares a signed ﬁeld intervention from
the ﬁrst native current, reads the ﬁeld with a binary Weyl instrument and feeds the bit back, with
inherited continuum bounds transferred to the complete history law. A source-compiled two-site
ﬁeld apparatus builds its entangled preparation from blank registers, evolves the ﬁeld and reads it
with a quantum pointer, all from emitted twelve-port control words; at the executed sixteen-level
cutoﬀ an exact rational interval places the ideal plus probability in [0. 51495, 0. 51789], excluding a
fair detector.
One interacting process. A common execution puts quantum electric ﬂux, a charged fermion
pair and two neutral scalar sites on one work register. Native source words implement their non-
commuting evolution, the matter and scalar pointer instruments, and feedback controlled by a
committed matter bit; the ﬁnal scalar bit controls the next native repair. A smooth extension
retains four quantum reads on the same disturbed state, compiles the C and E record laws around
it, and actively protects a logical qubit with record-controlled recovery. Its complete conditional na-
tive/work/reference outputs are retained. In the original sharp protocol, for zero native preparation
current the scalar feedback contrast of the same declared action has the certiﬁed inﬁnite-occupation
enclosure [−0. 024045670143, −0. 024043450955], including the original free vacuum, unbounded in-
teracting evolution and the actual coordinate half-line detector [ 71]. Section 4.3 develops what this
process says about measurement.
T wo observers, one law. Observer A reads ports (0, 1) and observer B ports (0, 5) of one
carrier, so the two regions overlap at port zero. Each receives its initial bits and every incident
local current, including zero currents, and its clock ticks whenever an attempted repair meets
its region while remote repairs continue on the same source. After the ﬁrst incident a native
feedback repair is chosen from the reconstructed occupation of port zero. Both observers decode
7

## Page 8

one continuing source, so their agreement at the shared port instantiates the overlap-agreement
equation with identity transport; rechartings between observer charts are not exercised. A direct
execution through the native endpoint branches and an independent recursion in a derived eﬀective
transfer representation agree exactly on all 5, 768 output words, including timeouts and terminal
populations; at every positive-weight preﬁx each observer’s decoded ﬁelds obey native continuity,
and the two values at the shared port agree [ 76]. The quantum extension checks the native remote
resolvent on all 4, 356 matrix units with coherent preparations and untouched references. A compiled
version replaces its ideal preparation, intervention and pointer by pulses emitted from the source
compiler: exact rational propagation of all 512 code addresses keeps every one of the 1, 656 clock,
timeout and abort outputs within 10−11 of the ideal history, at a cost of at most 51 pulses and ﬁve
native repair attempts [ 77].
Local clocks. A local event clock counts the attempted repairs that meet its region, or the
actual occupation changes at a port, and its stopped instrument sums the unresolved wait and
keeps every ﬁnite timeout. Two such clocks on joined carriers share one continuing history with
an observer-chosen ﬁeld. C and E have the same completed clock law there and diﬀer in the ﬁeld:
after three native attempts they predict ﬁeld-plus probabilities 35/ 54 and 14/ 27 under the stated
probe. Without feedback, over each two-tick cycle of one clock the other ticks twice on average with
variance 16, so equal mean rates do not imply equal readings, and converting between the clocks
through the joint event instrument preserves that whole count distribution [ 78]. A separate active
clock lets each incident event, including a zero-current one, trigger one compiled modular step whose
generator is the native three-state transfer, and closed histories derive a faithful thermal reference
for that generator. Four movement-bit covariances on the complete attempt tape recover its two
nontrivial frequencies and predict its coherent quadratures with no frequency ﬁt [ 79]. Attempt
counts, local event counts and modular parameters are distinct parameters, and none is identiﬁed
with seconds.
Identifying the repair law. Nine source-executed scalar measurements identify all nine real
coordinates of the record-law family and with them the coarse deﬁnite-current instrument on all 78
states, with arbitrary spectator and reference coherences. A collision test separates an admissible
law that agrees with E on the whole binary shell. The calibrated instrument predicts complete
two-event feedback histories through the same observer runtime, including quantum references
and checkpoint continuation, and an explicit bound carries the probability uncertainty to their
joint clock, current and interference records [ 80]. The probes are synthetic, so this identiﬁes an
instrument within the declared family and selects no law for nature.
F rom one carrier to a screen. One carrier suﬃces for the local results above; joined carriers
test larger regions, distant interventions and reﬁnement. Federations of 20, 80 and 320 twelve-port
carriers preserve complete source records, untouched references and cross-carrier coherence probes,
and local dependency bounds control the whole local quantum and record output under local
feedback. On the actual 15, 360- and 245, 760-port screens, matched radius-nine initial restrictions
give complete root-observer output distance below 3. 82×10−4 at auxiliary per-seam parameter 1/ 4,
and below 6. 656 × 10−9 after four local incident attempts with no auxiliary time coordinate [ 81].
On those matched screens a logical qubit with an untouched reference passes through two native
incident ticks and local record-controlled feedback while the surrounding source keeps density 1/ 3.
A degree-two operator calculation retains every quantum output and timeout, and each full-screen
output lies within 7. 779 × 10−10 of the common 222-port computation, uniformly over references.
8

## Page 9

C and E give identical classical records in this protocol and diﬀerent reference coherences [ 82].
These executions are ﬁnite, and their apparatus, services, schedules, ﬁeld action and prepa-
rations are declared; their time parameters are attempt and event counts. They discharge ﬁnite
execution, continuation and comparison between observers under those contracts. Autonomous
unbounded observers, an interacting spatial reﬁnement, physical clock calibration and responsive
gravity are not constructed (Section 5).
3 What observers agree on
Two observers who share a boundary can disagree about what it shows. If repair stopped in
diﬀerent places depending on who acted ﬁrst, the boundary would carry no fact of the matter. If
it always stops in the same place, the ﬁnal record is a public fact that no single observer owns.
This section states when the second case holds, and then shows that completed records carry the
familiar quantum rules.
A completed ﬁnite conﬁguration x is a ﬁxed point of the named repair-and-agreement map,
Rcons(x) = x.
This map is distinct from the universe-level closure operator of Section 6. Fixed-point existence,
observable determination and schedule independence are separate questions, and the following
results isolate them [ 4, 6].
3.1 Observable normal forms
Deﬁnition 3.1 (Observable quotient system) . An observable quotient system is a tuple S =
(Q, C, B, B ), where Q is a set of conﬁgurations, C ⊆ Q is the consistent subset, B is a set of
protected records, and B : Q → B is the record map. For b ∈ B , write Cb = C ∩ B−1(b).
The ﬁber Cb is unrealizable, reconstructing, or ambiguous according as its cardinality is zero,
one, or greater than one. This trichotomy avoids assigning a state to a record that has either no
consistent extension or several observationally indistinguishable extensions.
Theorem 3.2 (Canonical observable normalizer) . The following statements are equivalent:
(i) B|C is injective;
(ii) there is a unique map N : Q → C ⊔ {⊥} that preserves B on its C-valued outputs, ﬁxes every
point of C, is constant on each B-ﬁber, and returns ⊥ exactly on ﬁbers with no consistent
extension;
If B(Q) ⊆ B(C), the map is a total idempotent retraction Q → C.
Proof. Injectivity makes every nonempty Cb a singleton, which deﬁnes N and forces all its properties
ﬁber by ﬁber. Conversely, if c, c ′ ∈ C have B(c) = B(c′), the ﬁxed-point and ﬁber clauses give
c = N (c) = N (c′) = c′.
The normalizer is deﬁned by observable ﬁbers, independently of any repair schedule; a repair re-
lation implements it when the relation preserves B, reaches the consistent set and satisﬁes a liveness
condition. Approximate versions control the normalizer where exactness fails, through an optimal
two-output distance bound and summable naturality defects along reﬁnement towers. Canonicity
9

## Page 10

carries no promise of cheapness: for succinct Boolean boundary systems, deciding whether a pro-
tected record has a consistent extension is NP-complete, and deciding injectivity of the record map
is coNP-complete [ 4].
For the canonical single-site repair, every adaptive run is eventually constant because each
nonstuttering attempt strictly lowers the mismatch count; pathwise weak fairness makes the stable
state a normal form, and completeness with conﬂuence sends every weak-fair run to the same public
ﬁxed point. These results select no scheduler and provide no stochastic hitting law, rate, physical
clock, or reﬁnement theorem. Attempts obey a sharp capacity law. A scheduler can delay the
ﬁrst genuine repair for any prescribed number of attempts. When every reducible record sees a
genuine change among the next q + 1 attempts, a bounded-waste upper horizon of (q + 1) times
the initial mismatch count holds, and an independent-defect family attains the work-conserving
case exactly at q = 0. The input-independent compiler produces a ﬁxed federation with immutable
input ports; genuine node repairs decrease a dependency rank, and pathwise weak fairness gives
stable consensus and the correct output. Emission-order round robin ﬁnishes within its ﬁrst n
attempts, and the single-consumer structure bounds every accepted path by n(n + 1)/ 2, with an
emitted family attaining Θ( n2). General well-formed circuits instead admit 2n − 1 accepted steps,
attained by a fanout chain, where n counts the emitted nodes. These statements concern canonical
node repairs: the weak patch-frame relation can stutter fairly forever, and no physical scheduler,
clock, energy or hardware cost follows from these operation counts [ 4, 6].
3.2 Public records from local repair
For a prepared batch of support-closed transactions, join τ and σ in the conﬂict graph when
Wτ ∩ (Rσ ∪ Wσ) ̸= ? or Wσ ∩ (Rτ ∪ Wτ) ̸= ?, and iterate to a ﬁxed point: compute semantic
support inside every aggregate, rebuild the conﬂict graph on the expanded read and write sets,
merge every connected component, and repeat until the partition and all aggregate supports are
unchanged.
Proposition 3.3 (Transactional local diamond). Let an accepted repair be an aggregate transaction
with read set Rτ, write set Wτ, a read snapshot, and a snapshot-determined payload. Assume:
(i) every read set contains the semantic dependency closure of its write set, and every aﬀected
acceptance functional is revalidated at commit;
(ii) the terminal closure above exists, and each of its connected conﬂict components has one coherent
canonical aggregate payload;
(iii) a prepared component whose snapshot remains valid survives commits of independent compo-
nents; and
(iv) accepted commits preserve the protected record and strictly decrease the integer descent func-
tional.
Then every one-step quotient peak t ← −s − →u has a quotient state v with t − →v ← −u.
Proof. Two diﬀerent ﬁrst steps cannot originate in one conﬂict component, since that component
has a unique aggregate payload, so they originate in distinct components of the terminal closure.
Rebuilding the graph after every aggregate support expansion makes their ﬁnal read and write sets
disjoint, so neither commit changes the other’s read snapshot, and semantic dependency closure
ensures neither changes an acceptance functional consumed by the other. Revalidation and the
10

## Page 11

surviving component rule make both second commits legal, and with snapshot-determined pay-
loads and disjoint writes the two application orders agree; protected-record preservation carries the
equality to the physical quotient.
Theorem 3.4 (Consensus normal form) . Suppose the accepted repair relation satisﬁes the hypothe-
ses of Proposition 3.3, so that it has the one-step diamond property, strictly decreases the quadratic
descent functional V of Section 2, and is repair complete, meaning that its normal forms are exactly
the consistent states. Every maximal asynchronous repair schedule from one initial state then ter-
minates at the same consistent quotient normal form. The induced global repair map is idempotent,
ﬁxes consistent states, and preserves the protected record.
Proof. Quadratic descent makes the accepted relation terminating. Proposition 3.3 makes it lo-
cally conﬂuent, so Newman’s lemma [ 41] makes it conﬂuent. A terminating conﬂuent relation has a
unique normal form below each source, and repair completeness puts that normal form in the con-
sistent set. Applying the normalizer a second time does nothing, and protected-record preservation
holds along every accepted edge.
Atomic commits, disjoint write sets, or termination alone do not imply the local diamond; coun-
termodels omitting semantic dependency closure, canonical aggregation or the surviving component
rule are retained with the ﬁnite receipt. This is the precise sense in which a public fact is what sur-
vives every authorized comparison: the public object is the protected quotient record, independent
of the order in which observers repaired.
The law runs at scale. The public run uses the level-six cell rung of the geodesic icosahedral
reﬁnement tower: 81, 920 exact twelve-port carriers with one declared gluing of ports to neighboring
cells, 2, 580, 480 seams in all, and the seam-mean retraction as repair primitive. Sixteen shuﬄed
schedules from one seeded load terminate at the balanced class with no accepted move raising
V , and all sixteen return one quotient hash equal to the component multiset ﬁxed by the initial
totals. A standalone veriﬁer that imports no simulator code rebuilds the component multisets
from the primitive arrays, replays one schedule from its seed and recomputes every hash [ 2]. The
level-ten rung, with 20, 971, 520 carriers and 660, 602, 880 seams, terminates all sixteen schedules at
the balanced class in 84 to 100 sweeps with one quotient hash ﬁxed by the loads. Across the rungs
from ﬁve to nine, one initial load per rung gives an excess-energy contraction between 0. 81 and
0. 83 per sweep, and the sweep counts to the balanced class grow slowly with the carrier count,
from 51 to 79 at level six to 84 to 100 at level ten; these ﬁnite readings prove no all-size mixing
or termination-time bound. On the sixteen level-six schedules the number of unit transfers to the
balanced class lies between 924, 832 and 925, 708, about 1. 25 times half the initial ℓ1 deviation of the
loads from their mean. The ﬁrst-settling conﬁguration retains spatial information that its public
component multiset does not encode: at level ten the initial-to-terminal amplitude responses of
equal-count ancestral groups of sizes 1, 024, 4, 096 and 65, 536 are 0. 90758, 0. 95232 and 0. 98983,
with conditional schedule-noise fractions 1. 241, 0. 594 and 0. 143 percent. Stationary continuation
adds a constructive ﬂuctuation result: for fair integer repair on a connected weighted graph with
independent edge clocks of total rate one the ﬁxed-occupancy law is reversible, centered loads have
lagged covariance κe−|t|LG/ 2Π in the edge-clock time t, with LG the weighted graph Laplacian,
κ = r(p − r)/ (p(p − 1)) for r raised occupancies on p ports and Π the projection oﬀ constants, and
ﬁxed linear readouts obey a martingale central limit theorem [ 7].
The same records let observers recover the speciﬁcation itself. At carrier scale, recovery from
event logs under the declared probe and operation grammar returns the twelve ports, thirty seams,
inverse pairing, automorphism and rotation groups of orders 120 and 60, mean rule, schedule, rank-
three Gram and Lie split 1+ 3 + 8. Re-instantiation preserves the invariant vector for two iterations,
11

## Page 12

and tetrahedral and octahedral controls recover their distinct speciﬁcations [ 2]. Complete federation
gluing is not identiﬁed by the retained routing logs: within the declared port-gluing interface a two-
edge switch preserves every carrier degree, occupied port and source traversal while changing the
carrier graph’s triangle count from 14, 000 to 13, 990, and native executions reproduce identical bytes
for both. The alternative is not asserted to belong to the canonical geometric family. Full-gluing
recovery requires further observations or a justiﬁed restriction of the candidate class.
3.3 Completed records and quantum identities
Once a compare, write and verify slice has completed, its accessible events are represented in
a ﬁnite-dimensional ∗-algebra with a normalized state. This algebra-state representation is an
explicit input; the consensus theorem identiﬁes which record is public.
Theorem 3.5 (Finite public-record identities) . Let PE be the projector of a completed public event
E and ρ the normalized state of the completed record surface. The event weight and nonzero-weight
update
Pr(E) = Tr(ρPE), ρ | E = PE ρ PE
Tr(ρPE) (3)
deﬁne a normalized probability measure on every orthogonal event partition and a normalized pos-
itive state on the post-event corner. For self-adjoint dichotomic observables A0, A 1 and B0, B 1 in
commuting record subalgebras, with A2
i = B2
j = I, the combination SCHSH = Tr[ρ{A0(B0 + B1) +
A1(B0 − B1)}] obeys |SCHSH| ≤ 2
√
2 [22, 23]. The bound is attained exactly on the committed slot-
split interface by the maximally entangled state with Pauli and quarter-rotated projections, while
every state diagonal in the record basis with diagonal observables valued in [−1, 1] obeys |SCHSH| ≤ 2.
Proof. Positivity of ρ and 0 ≤ PE ≤ I give 0 ≤ Tr(ρPE) ≤ 1, and ∑
k Pk = I normalizes every
partition; for nonzero weight PEρPE/ Tr(ρPE) is positive and normalized. For B = A0(B0 +
B1) + A1(B0 − B1) the commuting-subalgebra calculation gives B2 = 4 I − [A0, A 1][B0, B 1] [24],
so ∥B∥ ≤ 2
√
2. The saturating witness and the diagonal bound are exact ﬁnite calculations [ 5].
At the level of weights the Born form is derived. Every valuation on the eﬀects of the ﬁnite
surface that is nonnegative, normalized and additive on coexisting eﬀects equals E ↦→ Tr(ρE) for a
unique density operator ρ, in every ﬁnite dimension including two, where the projector-only Gleason
theorem fails, and the derivation consumes no continuity axiom [ 20, 21]. The projectors alone do not
select the update: an explicit swap-twisted instrument has the same eﬀects and diﬀerent updates.
Kraus form with induced eﬀect P together with invariance on the states certain of P is equivalent
to being the Lüders map [ 27, 5]. The same surface carries the copying obstruction: if one ancilla-
free isometry copies two sharp states from a common blank, preservation of inner products forces
their overlap z to satisfy z2 = z, so two distinct alternatives copied by one device are orthogonal
[25, 26]. Finite webs do not by themselves supply the valuation hypothesis: a source-produced
complex-tomographically-complete eﬀect and instrument web with full coexistent-eﬀect additivity
is the missing step [ 5].
4 Selected results
Six results are developed here, three as theorems with proofs, two as certiﬁed computations with
their declared inputs and one as a source limit theorem with a proof sketch. They were chosen
because the executable construction consumes them or because they carry the programme’s most
speciﬁc structural content. Section 4.8 lists the other results with their classes.
12

## Page 13

4.1 Three spatial directions
The carrier incidence ﬁxes the dimension of the carrier’s own position readback, which is the reason
the number three occurs in this framework without being declared.
Theorem 4.1 (Intrinsic three-dimensional carrier completion) . Let A be the twelve-port adjacency,
let M = I − (5I − A)/ 60 be the uniform scalar repair mean, and take the complete equally weighted
centered probe census. The normalized inﬁnite-response kernel converges to the intrinsic Gram
G = 4 Pslow, of rank three. The twelve integer port loads have antipodal-odd quotient Z12/ {x :
xp = x−p} ∼= Z6, and the thirty-seam boundary has image exactly the even-sum sublattice D6, with
one residual parity coset. Under the pullback of the response-selected Gram metric, both integer
modules embed densely in one abstract continuous three-dimensional Euclidean carrier. The sixty
proper carrier maps act faithfully and isometrically on the rank-three range and extend to that
completion, and cumulative records act by exact internal translations.
The mechanism is spectral. The icosahedral adjacency has eigenvalues 5,
√
5, −1 and −
√
5
with multiplicities 1, 3, 5 and 3, so the repair mean M has eigenvalue 1 − (5 −
√
5)/ 60 on a three-
dimensional eigenspace and smaller eigenvalues on everything else nonconstant. Repeated local
averaging forgets every nonuniform pattern except a three-dimensional one, and that surviving
pattern is the triplet of ( 1) carried by the port position vectors. The same number (5 −
√
5)/ 60
reappears as the quantum spectral gap of the native repair channel (Section 2.3).
The completion is taken after slow-response normalization and the inﬁnite-step limit; at ﬁnite
step count the kernel has rank eleven on the centered space and rank six on the signed antipodal
sector, so the order of limits matters. The metric is the response-selected Gram pullback, distinct
from the usual six-dimensional lattice metric on D6. Thirteen ﬁnite covariance and isometry table
identities are discharged in Lean with native_decide, so those subreceipts trust the native com-
piler in addition to the kernel and are not kernel-only proofs. Displacement is a record diﬀerence
and cumulative records translate the completion, so position is the carrier’s own readback of its
accumulated records. In the executable observer the same rank-three readout attaches to all joint
histories of the interacting process, and a binary controller extension executes displacements span-
ning three independent directions [ 67]. Reading the completion as physical position consumes the
realization map (M1) of Section 5.2 [1, 7]. The rank-three readout belongs to one carrier’s response.
When carriers join into the reﬁning spherical screen of Section 4.4, the continuum form of their
repair history has spectral dimension two, the dimension of the screen, and bounded-energy families
lose every within-carrier component, this band included, at the present full-port energy scaling; a
three-dimensional bulk reading of the screen requires observation maps that the native energy does
not supply.
4.2 The gauge algebra and one generation
Let A be the icosahedral adjacency operator and J the antipodal involution on the twelve ports.
Exact incidence gives
10J = A3 − 4A2 − 5A + 10I, (4)
and the real port module splits as P12 ∼=A5 1 ⊕ 5 ⊕ 3 ⊕ 3′: the even coordinates carry 1 ⊕ 5 and
the odd coordinates the rank-three frame channel and its Galois-conjugate companion.
The theorem below uses two premises stated in the introduction and developed in Section 6.2.
The carrier layer of the speciﬁcation supplies a complete reversible response: an injective response
derivative D : P12 → u(H) whose image is commutator-closed with a positive trace pairing. The
13

## Page 14

no-outside rule applied to transformations requires every proper recharting of the carrier to be
implemented from inside that response; the library calls this the endogeneity clause of A2.
Theorem 4.2 (Gauge Lie-type forcing) . Let D : P12 → u(H) be the complete reversible response
derivative of the carrier, and put g = D(P12); let G0
D be the connected subgroup of the unitary group
of H generated by exp g. Suppose every proper carrier action a ∈ A5 has a projective implementer
satisfying
Ad[Ua] D(v) = D(a·v), [Ua] = [ gaca],
where ga ∈ G0
D and ca centralizes g pointwise. Then
g ∼= u(1) ⊕ su(3) ⊕ su(2).
The center is the unique trivial line. Up to exchanging 3 and 3′ by the outer automorphism of A5,
the su(2) ideal carries 3 and the su(3) ideal carries 3′ ⊕ 5.
Proof. Injectivity of D gives dim g = 12 , and the positive trace pairing makes the commutator-
closed image a compact Lie algebra. For every a ∈ A5 the centralizer factor acts trivially on g, so
the action is inner, and covariance through the injective map D gives a homomorphism from A5
to Int(g). The transitive proper-carrier action ﬁxes only the uniform port line, hence dim gA5 = 1.
Compactness gives g = z ⊕ [g, g], and inner automorphisms ﬁx the center pointwise, so dim z ≤ 1.
If z = 0 , the classiﬁcation of compact simple Lie algebras forces su(2)4 [58]; inner automorphisms
preserve each ideal, and on each three-dimensional ideal A5 acts trivially or faithfully, so its ﬁxed
space has dimension three or zero, since a ﬁnite rotation group ﬁxing an axis is cyclic. The ﬁxed-
space dimension of g would be a multiple of three, a contradiction. Hence dim z = 1, the semisimple
part has dimension eleven, and the classiﬁcation leaves only su(3) ⊕ su(2). The three-dimensional
ideal cannot carry the trivial action, so it carries 3 or 3′, and the eight-dimensional ideal carries
the complementary triplet together with 5.
The argument has a short intuitive form. Twelve ports give twelve independent ways to push the
carrier, and complete response turns them into a twelve-dimensional compact Lie algebra. Internal
implementation makes every recharting act by an inner automorphism, and inner automorphisms
leave the center pointwise ﬁxed, so the center lies in the directions the carrier symmetry ﬁxes. Only
the uniform push is ﬁxed. The center is therefore at most one-dimensional, the remaining eleven
dimensions must be semisimple, and 11 = 8 + 3 is the only compact split.
No ambient compact group or candidate Lie algebra occurs among the premises. The port count
twelve equals the dimension of the forced algebra, so the carrier ﬁxes that dimension by declaration;
the content of the theorem is the exclusion of the only other twelve-dimensional compact type
su(2)4 through compactness and the one-dimensional ﬁxed space, together with the forcing of the
unique center. The native record laws cannot supply D: every native seam expectation annihilates
a nonzero operator, so no composition of ﬁxed native repairs is invertible, and D belongs to the
reversible response layer of the carrier speciﬁcation. The reconstruction of D from ordered reversible
histories is consumed as a premise, and laboratory currents consume the realization map (M3) of
Section 5.2.
The forced algebra has an explicit ﬁnite realization. Choose from each antipodal axis one of
the certiﬁcate’s exact coordinate vectors ui ∈ Q(
√
5)3, each with ∥ui∥2 = 2 + φ for φ = (1 +
√
5)/ 2,
and put U = [u1 · · · u6]. For a real port ﬁeld f with antipodal involution J, write
bi = fi + fJi
2 , d i = fi − fJi
2 , c = 1
6
∑
i
bi, b 0 = b − c1,
14

## Page 15

so that the even coordinates b carry 1 ⊕ 5 and the odd coordinates d carry the rank-three frame
channel detected by U and its Galois-conjugate companion detected by σ(U ), where σ is the ﬁeld
automorphism
√
5 ↦→ −
√
5 applied entrywise. Let
Φ 0(b0) =
6∑
i=1
b0
i uiuT
i , ˆx y = x × y.
The conditional matrix witness uses the space
H = C3
E ⊕ C3
W ,
two copies of C3 carrying the exact rotations Rg and their Galois conjugates σ(Rg) deﬁned below,
and four nonzero rational coeﬃcients λ1, λ 5, λ 3, λ 3′, one on each source band. These are explicit
branch premises. The two three-dimensional response blocks match the two rank-three odd chan-
nels; they are not asserted to be a derived physical Hilbert space. Write vp for the signed coordinate
vector at port p. For g ∈ A5, let Rg ∈ SO(3) be the exact rotation satisfying Rgvp = vg(p) on the
oriented port frame, and deﬁne
Π( g) = diag( Rg, σ (Rg)).
The target A5 action on u(H) is conjugation by Π( g).
Theorem 4.3 (Conditional four-band ﬁnite port-response algebra) . On the declared twelve-port
carrier and supplied response representation, deﬁne
K(f ) = diag
(
λ3 ˆU d + i{λ1cI3 + λ5Φ 0(b0)},
λ3′ \σ(U )d
)
∈ u(H).
(5)
Then K is injective, its image is commutator-closed, and
K(g·f ) = Π( g)K(f )Π( g)∗.
The induced A5 action on im K is inner. Let K−1 : im K → P12 denote the inverse of K onto its
image. With the bracket pulled back from that image,
[x, y ]K = K−1([K(x), K (y)]mat) ,
one has
(P12, [ · , · ]K) ∼= u(3) ⊕ so(3) ∼= u(1) ⊕ su(3) ⊕ su(2). (6)
Its center is the uniform one-dimensional port line, its derived algebra has dimension eleven, and the
ﬁve-dimensional A5 band is noncentral. The map has a positive-deﬁnite invariant Hilbert–Schmidt
pullback.
Proof. Evaluating (4) on the four adjacency eigenspaces gives the antipode relation, and the source
protocol solves the common farthest-shell ﬁlter to give the signed response −J. The constant
coordinate maps to iRI3; the six rank-one axis matrices span the symmetric 3 × 3 matrices with∑
i uiuT
i = (5+
√
5)I3, so their sum-zero coeﬃcients map isomorphically onto the traceless symmetric
matrices under Φ 0; U is an isomorphism on the 3 band and annihilates 3′, with σ(U ) complementary.
Hence K has rank 12 and image exactly u(3) ⊕ so(3). Equivariance, positive deﬁniteness of the
invariant pairing, and the statement that Π( g) is the exponential of an element of im K are exact
calculations in Q(
√
5) over the full sixty-element domain, part of the proof by the exhaustive-
computation rule of Table 1 [1].
15

## Page 16

Boundary 4.4. The A5-module decomposition alone does not determine a Lie bracket: the same
module also carries the zero bracket, whose trivial response-generated group fails the A2 endogeneity
clause. The independent inputs to Theorem 4.3 are the carrier incidence and orientation, the
supplied response representation, and four nonzero band coeﬃcients. The block architecture and
coeﬃcients are declared. No measured particle, mass, or coupling value enters the exact veriﬁcation.
The theorem constructs the pulled-back bracket for this response map; it does not select K from
source histories or prove a uniqueness claim over all A5-equivariant brackets or ﬁnite carriers. The
abstract Lie type is forced independently by Theorem 4.2.
Theorem 4.5 (Finite matter image and exhaustive exterior-menu selection) . Let V = C3
−1/ 3 ⊕C2
1/ 2
be an explicitly supplied trace-balanced matter representation of the forced Lie algebra, with ˜G =
SU(3) × SU(2) × U(1) acting by (g3, g 2, z ) · (v3, v 2) = ( z−2g3v3, z 3g2v2). Its exterior module branches
as
Λ 2V ⊕ Λ 4V = Q ⊕ uc ⊕ ec ⊕ dc ⊕ L (7)
with Q = (3 , 2)1/ 6, uc = ( ¯3, 1)−2/ 3, ec = (1 , 1)1, dc = ( ¯3, 1)1/ 3 and L = (1 , 2)−1/ 2: ﬁfteen states,
all perturbative gauge and mixed gravitational anomalies vanishing, and an even number of weak
doublets. The common kernel of the action is Z6, so the maximal faithful image is
S(U (3) × U (2)) ∼= SU(3) × SU(2) × U(1)
Z6
. (8)
Among all 210 = 1 , 024 subsets of the ten nontrivial irreducible summands of Λ •V , imposing
nonemptiness, chirality and vanishing of the mixed gravitational– U(1), SU(3)2U(1), SU(2)2U(1)
and U(1)3 traces leaves exactly two rank-ﬁfteen masks, exchanged by charge conjugation.
Proof. The exterior-sum identity gives the ﬁve components and charges, and the anomaly sums
are the standard ones [ 62]: for example 6( 1
6 )3 + 3(− 2
3 )3 + 1 + 3( 1
3 )3 + 2(− 1
2 )3 = 0 , and with three
color copies of Q against one L the Witten parity is even [ 64]. The element (e2πi/ 3I3, −I2, e iπ/ 3)
acts trivially and generates a cyclic group of order six, and center-action enumeration excludes a
larger kernel. The scan evaluates the four traces on all 1, 024 masks; its only survivors are the two
exterior-parity masks.
The branching is the familiar 10 ⊕ 5 of SU(5) [61], and the quotient is a possible global form of
the Standard Model group [ 63]. One corollary concerns a channel the framework does not contain:
the forced adjoint (8, 1, 0) ⊕ (1, 3, 0) ⊕ (1, 1, 0) contains no mixed (3, 2, −5/ 6) generator, so the min-
imal simple grand-uniﬁed exchange channel is absent from the twelve-dimensional current algebra.
Separately, Doplicher–Roberts reconstruction supplies a compact sector group from realized bosonic
edge sectors [ 59, 60]; its equality with the ﬁnite response group is a separate commuting-square
premise. Physical matter identiﬁcation, family multiplicity, Yukawa couplings and masses consume
(M3) and (M5) [ 11].
4.3 Measurement inside one interacting process
OPH places measurement at the transition from a private repairable record to a protected public
normal form: a record is public when it has survived every authorized comparison and no repair
can change it, and event projectors, weights and updates act on the completed algebra-state surface
of Section 3.3. The executable observer puts this account inside one interacting quantum process
[66].
16

## Page 17

Quantum electric ﬂux, a charged fermion pair and two neutral scalar sites share one work
register, with operator Gauss law and source-executed instruments. The original sharp protocol
retains 1, 056 native/quantum histories. At d = 2 and 512 temporal steps its certiﬁed feedback
contrast is [−0. 033297300667839, −0. 030025660061775]; for zero native preparation current the
actual inﬁnite-coordinate half-line instrument gives [−0. 024045670143, −0. 024043450955] [71].
Outcome accuracy does not certify the output state. For that zero-current preparation, the
complete d = 2 sharp output has trace distance about 0. 455 from the inﬁnite history; every output
in the d = 256 occupation code has error above 0. 122. For the vacuum control the best code
error scales as d−1/ 4 [72]. A separate smooth Ramsey instrument preserves ﬁnite number moments.
A four-read continuation carries its disturbed state through two smooth reads, feedback and a
later gauge probe, retaining sixteen work branches. Stagewise moment and projected-Duhamel
estimates give complete O(1/d ) error at ﬁxed protocol with n(d) = d2 temporal steps; executed
d = 2 , 4 occupation bounds remain vacuous. This covers the later probe, not post-sharp clocks or
physical clock calibration [ 73].
Compiled C/E native repairs surround that same work history, acting on arbitrary joint inputs
of the native, work and reference registers. They retain full conditional quantum outputs, aborts
and padding. If r2 is the second smooth bit, Kj(h)
r the work branch, and e(0) = (0 , 5), e(1) = (0 , 1),
the joint Kraus map is
Mh,r,s = (Le(r2)
s L(0, 1)
h ) ⊗ Kj(h)
r .
Causal summation proves completeness. A retained native phase witness has C contrast (1 + q)/ 4,
q = Pr( r2 = 0 | j = 0) , and E contrast zero; source execution gives approximately 0. 3492 and
0. 3514 for C at d = 2, 4 [74].
Active memory uses the existing four-conﬁguration qubit V and compiled record-controlled
recovery RhLhV = √phV . Correcting after each native event, with ﬁnal seam (2, 3) for r2 = 0
and (0, 1) otherwise, preserves the logical/reference marginal even for entangled work inputs. For
independent logical/work preparations, the matched passive complete plus/minus output distance
is 1 − ¯q, ¯q = ( q−1 + 2q0 + q1)/ 4, qj = Pr( r2 = 0 | j), versus one with correction; a terminal phase
read consumes that coherence [ 75]. This protects an encoded qubit under speciﬁed controls, not
arbitrary lost native coherence.
Source selection has three primitive obligations: carrier architecture, observer interface with im-
plementation, and complete repair rule. A ﬁxed source-executed record read distinguishes coherent
implementations by 3/ 4 versus 1/ 4; committing ﬁrst yields 1/ 2 for both [ 98]. Eﬀective ﬁelds and
clocks must follow from the selected process; physical calibration, an interacting spatial reﬁnement
and empirical identiﬁcation require further evidence.
4.4 A continuum operator from one repair history
The executable carriers join into the geodesic spherical screens of the reﬁnement tower; the level- l
screen has Vl = 240 · 4l binary ports. The uniform canonical preparation ﬁxes the total load at
ql = Vl/ 3, every attempt selects one of the native seams uniformly and independently, and the fair
native repair acts on it. Observers read cell averages of the occupations against the normalized
area measure of the declared ﬂat polyhedral comparison surface, and χl = ql(Vl − ql)/ [Vl(Vl − 1)]
normalizes their ﬂuctuations. No continuum ﬁeld, metric or Hamiltonian enters the construction.
Proposition 4.6 (Native continuum operator) . Along one common subsequence of levels:
(C1) the native cell Green operators converge in operator norm to a positive, compact, injective
operator G on the mean-zero observation space, and A = G−1 is a densely deﬁned positive
self-adjoint operator;
17

## Page 18

(C2) with the constant mode restored, the form of A is a conservative, regular, strongly local Dirich-
let form with domain H 1, its energy lies between positive multiples of the gradient energy, its
positive eigenvalues obey n/ 49920 ≤ λn(A) ≤ 3n/ 2, and its heat-trace spectral dimension is
two;
(C3) for mean-zero test functions f the actual density readouts
Yl(τ; f ) =
∑
p fl(p) [n⌊1260·16lτ⌋(p) − 1/ 3]√Vlχl
at native attempt count ⌊1260 · 16lτ⌋ converge jointly at every ﬁnite set of values of τ to
a stationary reﬂection-positive Gaussian family with covariance ⟨f, e −|τ−u|Ag⟩, whose one-
particle semigroup is e−τ A;
(C4) the admitted coherent continuation converges strongly to cos(t
√
A), and with ¯A = A ⊕ 0 and
d the comparison distance,
∥1F e−τ ¯A1E∥ ≤ e−d(E,F )2/ (2τ),
so the cosine propagator vanishes between E and F while |t| <
√
2 d(E, F ).
Proof sketch. For conditional averaging Pk on the level-k cells, an elementary Poincaré inequality on
the actual subdivision gives ∥(I − Pk)u∥2 ≤ 624 · 4−k⟨u, B lu⟩ for every cell vector u and l ≥ k, where
Bl is the scaled native cell energy. It makes the Green family uniformly compact and upgrades weak
covariance convergence to norm convergence. Convergence of the inverses gives Mosco convergence
of the forms [ 93]. The full-port energy minimizer of a bounded-energy family has a within-carrier
part of order 2−l, so normal contractions of port values pass to the limit and the closed form is
Markovian and conservative [94]. Tent functions bound the eigenvalues from above and the Poincaré
inequality from below, which ﬁxes the heat-trace growth t−1. For (C3) each unread repair acts as a
fair identity or swap; the conditional proﬁle overlaps concentrate uniformly in attempt count, and
a conditional ﬁnite-population central limit theorem applies [ 95]. Clause (C4) is the Gaussian heat
bound together with its equivalence to ﬁnite propagation of the cosine family [ 96]. The complete
arguments and constants are given in [ 83, 84].
One positive operator therefore plays three roles for observers of one history: it is the inverse
covariance of the static ﬁeld, the generator of the ﬁeld’s decay in rescaled attempt count, and the
squared frequency of the admitted waves, whose locality gives them a ﬁnite speed. Two consecutive
long observation blocks of the same history also synthesize a continuous classical Gaussian wave
readout with covariance G cos((t − u)
√
A) in C([0, T ], H −2(S2)), with observation count taken to
inﬁnity before spatial reﬁnement. A repair whose execution is biased by ϵ toward a disclosed
coordinate proﬁle f has continuum mean derivative 2(I − e−τ A)f in its executed-repair clock, and
uniform likelihood bounds keep its response strictly positive at ϵ = 10 −6, with timeouts and coin-
service costs included [ 79]. With the same normalization the response and the ﬂuctuations obey a
ﬂuctuation-dissipation relation: for Cg,f (τ) = ⟨g, e −τ Af ⟩ the mean response of g to the bias toward
f is 2[Cg,f (0) − Cg,f (τ)], and its initial rate 2⟨g, Af ⟩ is set by the operator that generates the decay.
The parameters τ and t are rescaled attempt count and an admitted coherent-wave parameter, the
wave processor and its coherent services are declared, and the distance is that of the comparison
surface; none of them carries a laboratory calibration.
The proposition also ﬁxes what the screen is not. Its spectral dimension is that of the two-
dimensional screen. The eigenvalue count excludes a unitary identiﬁcation of A, up to scale and
a bounded mass shift, with a second-order elliptic operator on a smooth compact three-manifold,
18

## Page 19

and a boundary-trace test excludes a boundedly invertible energy identiﬁcation with the static
boundary action of a smooth compact three-dimensional Dirichlet bulk, whose Steklov eigenvalues
grow like √n [97]. A bulk reading of the screen therefore needs a radial observation algebra that
the native energy does not supply. The same min-max comparison ﬁxes the angular spectrum
of any sky ﬁeld read as a bounded congruent image of the native long-average ﬁeld: a precision
whose eigenvalues grow like ℓ2+θ is form-congruent with A only for θ = 0 , so such a ﬁeld has
the scale-invariant spectrum with ℓ(ℓ + 1)Cℓ bounded above and below, and the conditional tilt
branch of Section 7 needs an ingredient beyond this covariance. A speciﬁed random-length return
reference has modular generator arccos K, with K the symmetric native conﬁguration transfer,
and its modular phase magnitudes and cosine amplitudes match those of the waves. The full
conﬁguration reference approaches the normalized trace at ﬁxed modular parameter, so it identiﬁes
no quantum vacuum; its entropy deﬁcit is at most 2025/ (64V ) for V ports and every return cap. On
a complete base-face patch the stationary reduced entropy grows in proportion to the port count,
while the return reference’s mutual information per crossing seam vanishes, so neither carries a
nonzero universal coeﬃcient times the patch’s boundary-seam count; a reading of the whole screen
as the boundary of a three-dimensional region is not excluded, and that region map is a premise
[85].
Independent native long-block records nevertheless supply a radial product ﬁeld over the screen:
on a declared radial interval with Dirichlet ends its covariance is (−∂2
r + ¯A)−1, with spectral di-
mension three. The product wave action has coeﬃcient density C = µ diag(−1, a ), where µ is the
native measure and a the measurable spatial tensor, whose angular part complete native records
recover and whose radial coeﬃcient is one. The identity
g =
√
− det C C −1 = µ
√
det a diag(−1, a −1)
gives signature (−, +, +, +) almost everywhere [ 66].
Proposition 4.7 (Smooth (1 + 3) eﬀective descriptions). Smoothing the angular variables of µ and
a, with the radial Dirichlet boundary retained, converges in the norm-resolvent sense with spectral
gap (π/L )2. Consequently, for every ﬁxed ﬁnite admissible menu of observer readouts on bounded
wave times (thermal ﬁeld covariances, responses, ordered Weyl and Ramsey histories, detector and
reference outputs) and every ε > 0, one smooth static (1 + 3)-dimensional Lorentzian metric carries
a free scalar ﬁeld with the standard canonical and KMS normalization whose predictions for the
whole menu agree with the native ones within ε, plus the declared source and clock budgets.
Proof sketch. The native density relative to smooth sphere volume is positive and globally Lipschitz.
The unitary Uεf =
√
ρε/ρ f between the changing and ﬁxed L2 spaces converges on the form
domain; Rellich compactness, recovery sequences and the dual energy liminf give transported Mosco
convergence, which compact resolvents and self-adjointness upgrade to norm-resolvent convergence.
Thermal covariance and sine response then converge in operator norm, momentum covariance on
the form domain, and every ﬁxed ﬁnite history with them [ 66, 93].
The observers therefore have the eﬀective description that standard physics uses, a Lorentzian
ﬁeld theory in 1 + 3 dimensions, as a theorem about their own readouts. Its scope is stated with
it. The exact native measure, the radial pushforward of ﬂat face area, is not C1 across the original
face edges, so the proposition asserts no exact smooth native metric and no computable smoothing
rate. The carrier’s own records do supply a radius: read through the recovered rank-three frame,
the axial current counter converges from every binary initial state to three-dimensional Brownian
motion, and its norm has the radial generator 2
99 (∂2
r + 2
r ∂r) with angular term 2
99 r−2∆ S2, without
19

## Page 20

added radial sites. That radius is the radius of a retained record; it is not identiﬁed with physical
transport, and it does not select the radial architecture of the ﬁeld: complete adaptive histories
with angular instruments, native clock marks and feedback are identical across distinct radial
extensions [ 66]. The radial architecture and the active observer protocol therefore remain declared.
The metric is static and does not respond to stress. A quartic projection on the same product
has a source-prepared interacting Gibbs state, an active native event clock and an unrestricted-
occupation limit within its two modes, and the same smoothing preserves its ﬁnite histories; it is
no spatially reﬁning interacting continuum [ 66]. Uniqueness of the limit along the whole sequence,
an exact smooth native metric, responsive geometry and a spatially reﬁning interacting ﬁeld are
open.
4.5 A causal precursor from records
The oriented conformal spherical support carries the classical isomorphisms
Conf +(S2) ∼= PSL(2, C) ∼= SO+(3, 1), H 3 ∼= SO+(3, 1)/ SO(3), (9)
which identify the three-dimensional space of future unit timelike directions [ 47]. Authenticated
records separately supply events and their order. Every quotient-visible register version carries the
identiﬁer of its last semantic writer, each commit certiﬁes the versions its acceptance depends on,
and the informational order is the transitive closure of authenticated read-from parenthood, an
authenticated analogue of happened-before [ 37]. As an abstract ﬁnite poset it satisﬁes the order-
theoretic causal-set axioms, and every ﬁnite causal set is the generated order of some authenticated
log [ 1].
Theorem 4.8 (Source-derived ﬁnite (1+3) causal precursor) . Let L be an authenticated ﬁnite
semantic-event log with generated strict order ≺L and canonical source height hL, zero on roots and
one plus the maximum parent height otherwise. Then hL strictly increases along ≺L and equals the
maximum length of an authenticated parent chain ending at each event. Let Vsrc be the rank-three
source Gram quotient with positive form gsrc, and put Wsrc = R ⊕ Vsrc with Q(t, x ) = t2 − gsrc(x, x ).
(F1) Wsrc has dimension 1 + 3 = 4 and Q has one positive and three negative directions;
(F2) source-unit directions, S2 and future-null rays are exactly equivalent;
(F3) a placement F (e) = ( τ hL(e), x (e)) sends every generated precedence into the future cone
whenever each parent edge obeys gsrc(x(f ) − x(e), x (f ) − x(e)) ≤ τ2(hL(f ) − hL(e))2;
(F4) every ﬁnite log admits an injective placement of this form that sends generated ancestry into
the future cone and makes distinct same-height events spacelike; and
(F5) for a placement obeying the edge condition of (F3), if the spatial readback separates each
equal-height layer and every unrelated pair with hL(e) < h L(f ) is spacelike, then e ⪯L f holds
exactly when F (f ) − F (e) is future causal, so F is an exact order embedding preserving every
causal interval.
Proof. Dimension is additive and positivity of gsrc gives the inertia; induction over parents bounds
every ending chain by the height and constructs an attaining chain; the edge inequality and transi-
tivity of cone and ancestry give (F3); an enumeration along one axis with τ = n + 1, n the number
of events, gives (F4); and the two geometric checks exclude every unsupported cone comparison in
(F5). The ﬁnite statements are formalized [ 1, 9].
20

## Page 21

Read informally, the theorem combines two things observers have. Who read whose write is
a partial order, and the longest chain of reads ending at an event gives it a height. The rank-
three carrier gives three spatial directions and, with a real axis, a light cone. An event placement
that respects both turns the provenance order into a causal order in a four-dimensional Lorentzian
carrier.
The ﬁrst two clauses construct the carrier and signature from an independent real axis and the
rank-three source carrier, without assuming a rank-four chart; source height enters only through
placements, and the scaled height is no operational clock. A physical manifold requires a source-
causal continuum certiﬁcate: one reﬁnement family supplying physical event identiﬁcation, source-
selected order and cone embeddings, reﬁnement maps, a calibrated count density, agreement across
independent dimension and manifoldlikeness tests against Poisson-sprinkled controls, and conver-
gence to a distinguishing time-oriented Lorentzian geometry [ 28, 29, 30].
Proposition 4.9 (Composition of the continuum certiﬁcate) . For a source-causal continuum cer-
tiﬁcate, causal order ﬁxes the conformal Lorentzian geometry of the admitted distinguishing limit
and the calibrated count measure ﬁxes its conformal scale [ 35, 36]; the local model is (1 + 3), so a
limit satisfying the dimension and topology clauses of the certiﬁcate is an eﬀective four-dimensional
time-oriented Lorentzian event manifold. If on the same family a causal-set d’Alembertian or
scalar-curvature estimator converges as an independent curvature diagnostic [ 33, 34], an explicit
tensor-curvature reconstruction converges to the Einstein tensor or the small-ball null-balance route
identiﬁes the limiting geometry tensor with Gab, and stress and Ward–Bianchi data converge with
calibrated coupling, then the smooth limit is Gab + Λ gab = 8πG Tab.
A source-record population realizes the ﬂat case.
Theorem 4.10 (A source-record causal and count limit) . In the rank-three conservative repair
carrier, let φ = (1 +
√
5)/ 2, L = 2 / √φ + 2 and q = Fn the n-th Fibonacci number, n ≥ 4. The
population Sq = L{bφ − ⌊bφ⌋ : 0 ≤ b < q }3 ⊂ [0, L ]3 consists of q3 distinct conservative source-
record images. Retain it at each layer and let every next-layer event read all preceding-layer sites
within aq = L/ √q, with model duration ∆ q = aq/c for a declared speed c. As q → ∞ the order
converges away from null separation to the speed- c causal order, and with ρq = q3/ (L3∆ q), for ﬁxed
interior timelike diamonds,
|Iq|
ρq
− →Vol(I), 2Cq
|Iq|(|Iq| − 1) − → 1
10 ,
where Cq counts strict causally ordered pairs; for vertical duration T , Vol(I) = πc3T 4/ 24.
Corollary 4.11 (Proper-time ratios from record counts) . If Iq, J q approach interior timelike di-
amonds of proper durations τI , τ J > 0, the observer readout Rq(I; J) = ( |Iq|/ |Jq|)1/ 4 → τI /τ J .
Within positive volume-only readouts additive on collinear inertial intervals, the fourth root is unique
up to units.
At ﬁnite regulator the population is produced by the carriers themselves: each holds its record
as port loads, reads its position through the rank-three response and decides its neighbors from
that readback, and the precedence generated from their authenticated reads equals the layered
order through q = 89, that is 704, 969 carriers reading one another over ten rounds. On the largest
interior diamond the inverted Myrheim–Meyer dimension reads 2. 83, 4. 38, 3. 55, 3. 84 and 3. 92 at
q = 13 , 21, 34, 55, 89 [31, 32]; the outermost central diamonds, which the cube clips and which are
therefore not interior readings, read 4. 15, 3. 61, 4. 10, 4. 09 and 4. 07, and the readings at q = 5 and
21

## Page 22

q = 8 lie below the regime of the theorem. The ordering fraction of the central diamond reads 0. 088,
0. 138, 0. 092, 0. 093 and 0. 094 against the proved ﬂat value 1/ 10, and the count clock reads 1. 71,
2. 09, 1. 91 and 2. 05 at q = 13 , 34, 55, 89 against the model-time ratio 2. Covariance requirements
select the cone: a closed convex pointed displacement cone preserved by the irreducible icosahedral
rotations and by boosts along one axis is the future or past cone of t2 − ∥x∥2. The read law and
model clock are declared, and the ordinal height of a history with self-waits is no proper-time
instrument [ 9].
The executable carrier has its own ﬁnite causal bound. For a ﬁxed seam schedule and shared
local coins, dependence crosses at most one seam per ordered repair event, and retained signal
witnesses attain the bound; this is a speed in seams per native event. Remote repairs can carry
a diﬀerence to a receiver before its ﬁrst local tick, so the bound is not one seam per receiver
tick [ 86, 67]. Complete endpoint records give a forward-cone placement in the registered response
geometry, and a four-event native execution has a Lean instance proving that implication and the
failure of its converse [ 1].
4.6 The Einstein equation from a realized null balance
Theorem 4.12 (Null tomography and metric ambiguity) . Let V be four-dimensional with Lorentz
form η of signature (−, +, +, +), the opposite convention from Q in Section 4.5; nothing below
depends on the choice. If a symmetric form X satisﬁes X(k, k ) = 0 for every null vector k, then
X = ϕη. Null-null values therefore determine a symmetric tensor modulo the metric line, and
nine explicit null directions, the six axis rays (1, ±ei) and (1, s, s, ±s), (1, s, −s, s ) with s =
√
3/ 3,
reconstruct the quotient Sym2(V ∗)/ Rη with design determinant 8192/ 27 in the certiﬁcate’s basis
and a decoder obeying ∥x∥∞ ≤ (2 +
√
3)∥q∥∞.
Proof. On future null rays (1, n ) the identity X00 + 2X0ini + Xijninj = 0 on the unit sphere forces
X0i = 0 , Xij = ϕδij and X00 = −ϕ. The determinant and decoder bound are exact calculations,
and the design, injectivity and ambiguity are formalized in Lean [ 1, 9].
Theorem 4.13 (Tensor completion of a realized null balance) . Let (M, g ) be a connected, time-
oriented, C3 Lorentzian four-manifold and Tab a symmetric C1 tensor with ∇aTab = 0. Suppose a
common realized tower supplies κ = 8πGN > 0 with
G(g)
ab kakb = κTabkakb (10)
for every null vector k at every point. Then there is a constant Λ with
G(g)
ab + Λ gab = 8πGN Tab, (11)
and a reference event at which the diﬀerence equals −Λ 0gab ﬁxes Λ = Λ 0.
Proof. Theorem 4.12 applied pointwise to G(g) − κT gives G(g)
ab − κTab = ϕgab. The contracted
Bianchi identity, the Ward premise and metric compatibility give ∇bϕ = 0, connectedness makes ϕ
constant, and Λ = −ϕ.
The structure of the argument is simple. Light rays cannot see a stress term proportional
to the metric, so balance between curvature and stress on every null direction ﬁxes Einstein’s
equation up to such a term, and the Bianchi identity with stress conservation makes its coeﬃcient
constant. That constant is the cosmological constant. The coeﬃcients of entanglement-equilibrium
arguments [ 54] enter through one machine-checked identity: scalars with δSbulk = 8π2ℓ4
15 t, δA =
22

## Page 23

− 4πℓ4
15 f and δSbulk + δA/ (4GN ) = 0 obey f = 8 πGN t. For OPH to supply ( 10), one common
reﬁnement tower must realize (G1) the eﬀective manifold of Proposition 4.9; (G2) geometrically
normalized cap modular ﬂow whose directional charges deﬁne one stress tensor; (G3) its Ward
identity, independently of that reconstruction; (G4) the edge-center entropy split, the normalization
δ⟨ZL⟩ = δA/ (4GN ), the modular ﬁrst law and generalized-entropy stationarity; (G5) a small-ball
family with controlled remainders and enough observer directions; and (G6) universal coupling, a
vacuum reference and a scale calibration ﬁxing κ = 8πGN . The ﬁnite layer contains null tomography,
the edge-entropy split and the coeﬃcient identity, and the machine-checked composition proves the
tensor step from an explicit null-balance premise; the bridge between them is (M4).
4.7 The dark sector: non-luminous charge, the deep galaxy law, and the vacuum
coordinate
Under supplied universal-coupling, physical-stress and source-localization premises, sourcing gravity
by modular charge classiﬁes any geometric excess over the luminous term as non-luminous modular
charge, with anomalous modular energy on overlap collars as one candidate carrier [ 14]. Condi-
tional on two named premises the deep-regime proﬁle is characterized rather than supplied: an
enclosed-mass law that is degree-one scale covariant and composes independent sources in quadra-
ture carries a unique constant a0 > 0 with MA(r) = r
√
Mba0/G , which forces ﬂat rotation curves
and gives v4 = GMba0, the baryonic Tully–Fisher relation, each premise load-bearing by an explicit
counterexample. A per-cut model gives a0 = Gn2c, with only the product n2c identiﬁable, and the
value of a0 is ﬁxed by no source law. Reading the dark-energy density as ρDE = κ/N at constant
κ gives w = −1 + 1
3 d ln N/ d ln a, so ﬁxed capacity gives the exact point (w0, w a) = ( −1, 0) and
nondecreasing capacity gives w ≥ −1; exclusion of that point fails the ﬁxed branch, while agreement
supports nothing, since it is also the Λ CDM null [ 15].
4.8 The wider inventory
Table 2 lists the structures of observed physics that the reconstruction returns, with the class of
each result and where it is proved or stated. A ﬁnite theorem is a theorem about the declared
ﬁnite objects, a reconstruction implication carries the physical antecedents displayed with it, and
a diagnostic consumes measured input and predicts nothing.
Table 2: Reproduced structures of observed physics, the result that
returns each one, its claim class, and its location.
Structure What the reconstruction returns Class Where
An observer that
records and acts
A ﬁnite observer executes repair, retains its
records and chooses later interactions from them;
feedback changes the outcome law against
disconnected and shuﬄed controls
certiﬁed ﬁnite
computation
§ 2.4
Measurement inside an
interacting ﬁeld process
Flux, charged matter and a scalar ﬁeld on one
register with complete instruments and feedback;
a certiﬁed nonzero inﬁnite-occupation response
certiﬁed ﬁnite
computation
§ 4.3
Agreement between
overlapping observers
Two observers sharing a port read one complete
law of ﬁelds, local clocks, feedback and timeouts;
native and eﬀective descriptions agree exactly,
with a quantum extension and a compiled
apparatus within 10−11
certiﬁed ﬁnite
computation
§ 2.4
23

## Page 24

Table 2 continued
Structure What the reconstruction returns Class Where
Clocks made of events Local event clocks with complete stopped
instruments; conversion between two clocks keeps
their whole count distribution; an active clock
whose frequencies are recovered from record
covariances
certiﬁed ﬁnite
computation
§ 2.4
A local ﬁeld operator on
the screen
One repair history gives a regular, strongly local
Dirichlet form of spectral dimension two,
Gaussian density laws e−|τ−u|A and admitted
waves cos(t
√
A) of ﬁnite speed
source limit
theorem
§ 4.4
An eﬀective Lorentzian
ﬁeld theory
On a declared radial product, a Lorentz-signature
wave density and one smooth static (1 + 3)
metric whose free scalar reproduces every ﬁnite
menu of observer readouts to any tolerance
source limit
theorem
§ 4.4
A ﬁnite signal speed Dependence crosses at most one seam per
ordered native repair event, and retained
witnesses attain the bound
ﬁnite theorem § 4.5
Schedule-independent
public facts
Every maximal asynchronous repair schedule
reaches the same protected normal form under
the transactional local diamond
ﬁnite theorem § 3
Probability weights on
outcomes
Every nonnegative normalized valuation additive
on coexisting eﬀects is the trace against a unique
density operator, in every ﬁnite dimension, with
no continuity axiom
ﬁnite theorem § 3.3
State update after an
outcome
A swap-twisted instrument has the same eﬀects,
so the eﬀect table does not select the update;
invariance on the outcome-certain states selects
the Lüders map
ﬁnite theorem § 3.3
The correlation ceiling
2
√
2
Tsirelson bound on commuting record
subalgebras, attained exactly on a declared
entangled witness, with record-diagonal states
conﬁned to 2
ﬁnite theorem § 3.3
Impossibility of copying
an unknown state
One ancilla-free isometry copying two sharp
states from a common blank forces their overlap
to 0 or 1
ﬁnite theorem § 3.3
A three-dimensional
position carrier
The normalized twelve-port repair response has
an intrinsic rank-three Gram quotient into which
the record modules embed densely
ﬁnite theorem § 4.1
Lorentz kinematics and
the velocity space
Conf +(S2) ∼= SO+(3, 1) with H 3 the space of
future unit timelike directions
classical
identiﬁcation
on the
supplied
spherical
support
§ 4.5
A source-derived 1 + 3
causal precursor
The real axis and the rank-three source quotient
give Lorentz inertia (1, 3) with the S2 null
boundary, and authenticated logs embed by
exact order embeddings
ﬁnite theorem
plus
reconstruction
implication
§ 4.5
Proper time from
counting
Record counts in source-record diamonds recover
proper-time ratios by the unique fourth-root
readout
ﬁnite theorem § 4.5
24

## Page 25

Table 2 continued
Structure What the reconstruction returns Class Where
The Einstein ﬁeld
equation with Λ
Tensor completion of a realized null balance
under Ward conservation, the Bianchi identity,
and one scale identiﬁcation
reconstruction
implication
§ 4.6
The Standard Model
gauge Lie type
Complete response with internally implemented
rechartings forces u(1) ⊕ su(2) ⊕ su(3) on twelve
ports
ﬁnite theorem § 4.2
One anomaly-free
generation of ﬁfteen
states
Exhaustive scan of the declared exterior menu
leaves one charge-conjugate pair of rank-ﬁfteen
chiral anomaly-free masks, with common kernel
Z6
ﬁnite theorem § 4.2
Absence of the minimal
grand-uniﬁed exchange
channel
The forced adjoint contains no X/Y generator ﬁnite theorem § 4.2
The relativistic mass
shell and internal energy
in inertia
E2 = m2 + |p|2 for the declared-mass momentum,
and two declared selection routes giving inertial
coeﬃcient m + E
ﬁnite theorem [ 8]
A nonlinear matter
continuum and an
interacting state space
First-order trajectory convergence for the real
quartic-wave sector of a supplied charged action,
and a unique self-adjoint closure of its declared
Hamiltonian
conditional
analytic
theorem
[8]
The arrow of time and
the four laws
Reference-preserving stochastic repair contracts
relative entropy; one conditional package gives
the four laws, ﬂuctuation relations and the
Landauer bound
ﬁnite theorem [ 8]
Entropy that scales with
a boundary
Collar decomposition with one-sided edge term∑
α pα log dα ; the sparse crossing-read functional
Γ-converges to perimeter; on the two-dimensional
source the native return reference has mutual
information at most log J + 2025/ (64V ), too
small to scale with the crossing seams of a
base-face patch
ﬁnite theorem;
conditional
analytic
theorem
[9, 85]
Coulomb and Maxwell
structure
Unique minimal-energy neutral Coulomb
solution, a local twenty-face curvature action
gauge invariant for conserved seam currents, and
a hopping charge conserved at a declared
nonzero step h
ﬁnite theorem [ 7]
Massless photon, gluon
and graviton
Three conditional structural statements with zero
hard mass parameters on their declared branches
reconstruction
implication
[11]
A positive Yang–Mills
mass gap
The continuum gap identiﬁed with the limiting
repair gap under named ﬁnite-rate, transfer and
reconstruction certiﬁcates
reconstruction
implication
[16]
The Koide relation Positive-chamber circulant identity with Q = 2/ 3
exactly at ρ/a = 1/
√
2, with a custody-bound
tau window
ﬁnite theorem;
prospectively
ﬁxed
conditional
test
[12]
The ﬁne-structure
constant
Interval-certiﬁed screen-grain roots 136. 994835
and 137. 035660 against 137. 035999
diagnostic § 6.5
The cosmological
constant
Capacity candidates 3. 2921 × 10122 and
3. 3001 × 10122 against 3. 3129 × 10122
diagnostic § 6.5
25

## Page 26

Table 2 continued
Structure What the reconstruction returns Class Where
Flat rotation curves and
the baryonic
Tully–Fisher relation
Under deep scale covariance and quadrature
composition the unique law gives v4 = GMba0;
a0 is not source-derived
conditional
characteriza-
tion theorem
§ 4.7
The dark-energy
equation of state
Fixed record capacity under ρDE = κ/N gives
(w0, w a) = ( −1, 0)
reconstruction
implication
§ 4.7
The de Sitter transfer
sign
Exact capacity identity with
∆ Smax
gen = log(1 − f ) < 0
ﬁnite theorem [ 13]
Isotropy of propagation
to angular rank ﬁve
The ﬁrst symmetry-allowed anisotropic rank of
the carrier class is six, with two frozen rays
inside one exact band
ﬁnite theorem;
frozen
prospective
branch
prediction
§ 7.1
Bell correlations beyond
local records
Public loophole-free records give S = 2. 42250
against the bound S ≤ 2 of the diagonal branch
read as a local response model
diagnostic § 7.3
An instrument
comparison ﬁxed before
data
One coherent event reads the record law’s
coherence multiplier g; C predicts 5/ 8 and 1/ 8, E
predicts 3/ 8, with a complete-count rejection rule
prospectively
ﬁxed
conditional
test
§ 7
5 The speciﬁcation and its scoreboard
5.1 The completion target
The programme has one target: an executable realization of the observer algebra whose compatible
inter-observer measurements reproduce the observed universe. Observers inside the construction
must make the measurements we make and must be able to use the eﬀective descriptions we use:
their joint statistics must admit a nontrivial 3 + 1-dimensional quantum ﬁeld theory as a controlled
eﬀective description, including the observed matter and interactions, and classical physics and
dynamical gravity under appropriate coarse-graining. No fundamental ﬁeld or spacetime has to be
reconstructed ﬁrst. Fields, particles and geometry can be eﬀective descriptions of what observers
measure and jointly interpret, valid in their regimes, without being primitive objects. Exact equality
of limiting observable laws is the mathematical target; ﬁnite executions and experiments are judged
with declared approximation and measurement errors.
Table 3 gives the acceptance criteria and the status of each. The third column lists the evidence
executed with retained receipts or proved on its stated domain; “supplied” means consumed as a
declared input of the construction that uses it. Only the ﬁrst obligation is discharged, under its
declared contract. The next three are not discharged, and the last has its decision rules ﬁxed
without independent data.
26

## Page 27

Table 3: Acceptance criteria of the completion target and their
status.
Obligation Acceptance evidence Executed or proved Supplied or not
constructed
Executable
observer algebra
Explicit state, operation,
preparation, measurement,
retained-record and feedback
maps, executed, with
accounted memory and
services
Native observer,
electromagnetic and
scalar detectors,
source-compiled ﬁeld
apparatus, the common
interacting process with a
four-read continuation,
compiled C/E
attachment and active
encoded memory;
overlapping observers
with local event clocks on
one carrier; a compiled
clock history; complete
identiﬁcation of the
record law; quantum
federations and a
coherent observer on
matched screens
Conserved totals,
densities, coherent
apparatus, schedules and
services are declared;
stationary preparation
within a ﬁxed-number
sector follows from a
ﬁnite native warm-up
with a derived error
bound; autonomous
unbounded observers and
physical memory
calibration are not
constructed
Shared
operational
meaning
Interpretation maps ﬁxed
independently of validation
outcomes, compatible across
observers and reﬁnement,
preserving intervention
statistics and quantum
correlations
Schedule-independent
public records
(Theorem 3.4); the
completed interacting
history under every fair
schedule; one complete
law for two overlapping
observers, exact between
native and eﬀective
descriptions, with a
quantum extension;
complete-output maps
between actual screen
levels
Interpretation maps for
physical quantities across
all observers and
reﬁnement levels are not
constructed
27

## Page 28

Table 3 continued
Obligation Acceptance evidence Executed or proved Supplied or not
constructed
Eﬀective
quantum-ﬁeld
description
A reﬁning family of observer
statistics whose errors
relative to a nontrivial
3 + 1-dimensional QFT tend
to zero, including observed
matter and interactions
One interacting register
with operator Gauss law;
inﬁnite-occupation and
complete-output
certiﬁcates for a two-site
action; free-ﬁeld
continuum theorems; a
native continuum
operator for Gaussian
density ﬁelds, their decay
and admitted ﬁnite-speed
waves on the screen
(Proposition 4.6); a
smooth static (1 + 3)
Lorentzian eﬀective
description of every ﬁnite
readout menu on a
declared radial product
(Proposition 4.7), with a
ﬁnite interacting Gibbs
projection and active
native clock on that
product
The interacting ﬁeld
action, vacuum and
coupling and the radial
product are supplied; the
native reference identiﬁes
no quantum vacuum; the
exact native metric is not
smooth in the ﬁxed atlas,
and an interacting
reﬁning family is not
constructed
Classical and
gravitational
descriptions
Compatible classical,
geometric, stress and clock
descriptions of the same
observations, reproducing
gravitational predictions
with controlled corrections
Rank-three displacement
readouts; a native causal
bound of one seam per
repair event; the ﬁnite
(1 + 3) precursor, ﬂat
count volume and tensor
completion of null
balance; a
two-dimensional local
continuum form on the
screen with a controlled
repair response
2(I − e−τ A)f , the entropy
bound log J + 2025/ (64V )
and a boundary-action
test against a
three-dimensional
Dirichlet bulk
Read law, model clock,
stress, region map and
scale premises are
supplied; a
source-selected radial
reading and responsive
geometry on the same
process are not
constructed
28

## Page 29

Table 3 continued
Obligation Acceptance evidence Executed or proved Supplied or not
constructed
Physical
predictions
Input ancestry, calibration
and decision rules ﬁxed
before independent data,
agreement with observed
statistics
The C/E protocol with a
sequential trial design
and the angular-rank-six
band are ﬁxed before
comparison; retrospective
replays of public Bell and
optical-clock records test
stated attachments of
native records, and the
conditional tilt branch is
compared with CMB
posterior chains
Physical encoding,
service calibration and
independent prospective
data are not supplied; the
retrospective comparisons
establish no common
physical observation
dictionary
5.2 The standing realization maps
A statement about nature needs one physical realization carrying compatible events, ﬁelds, action,
states and readouts across reﬁnement. Five maps name what such a realization must supply, and
each boundary statement in this paper cites them by label.
(M1) Position and scale. An identiﬁcation of the intrinsic carrier completion with physical po-
sition, with a physical length, coﬁnal overlap gluing between carriers, and a reﬁnement limit.
On the reﬁning spherical screens the native continuum operator has spectral dimension two on
the two-dimensional source, and bounded-energy families lose the within-carrier band there
(Section 4.4).
(M2) Clock and energy calibration. An identiﬁcation of the internal step with a laboratory
clock and of the modular ledger with a laboratory energy, in measured units; conﬁguration-
derived model duration does not select laboratory time or its units. Native event clocks, an
active clock and an inverse clock prediction supply joint clock tests in count parameters, and
a direct count-to-phase attachment to published optical ratios is rejected under its readout
assumptions (Section 7.3).
(M3) Laboratory current. An identiﬁcation of the ﬁnite port response with a physical gauge
current, with a physical matter action and a continuum operator limit.
(M4) One common physical tower. A single physical reﬁnement family realizing the event,
modular, stress, entropy, continuum, vacuum, and scale premises together, rather than several
independently declared islands.
(M5) A source-selected action. An action produced by the source rather than declared, with
its ﬁeld sector, coupled interaction, relative normalizations, and operational readout.
The executable observer turns these maps into construction work on one process. Its retained
histories carry a rank-three displacement readout, local event clocks, modular clock probes and
committed records on a single register, and its joined screens carry one continuum operator for
ﬁelds, decay and waves, with a smooth eﬀective Lorentzian description on a declared radial product.
What such a realization lacks is an interacting reﬁning family whose records, clocks, currents and
29

## Page 30

stress readouts converge together in physical units. A ﬁnite theorem that consumes none of the
ﬁve maps stands on its own domain, and no result in this paper discharges any of them; Section 7.3
tests two simple attachments of native records to published data.
5.3 Open connections
These connections can be built in parallel, and each concerns the same process:
(i) extend the native continuum operator from passive ﬁelds and admitted waves to the full in-
teracting construction, with a quantum reference state and an eﬀective Lorentzian description
that is interacting and responsive, while retaining records and bounding numerical and physical
service errors;
(ii) attach interacting matter, a source-selected radial observation algebra, responsive geometry
and physical clock calibration to the same state, event law and retained records;
(iii) extend ﬁnite composition and active encoded memory to autonomous operation with accounted
storage, reset and service contracts;
(iv) realize the ﬁxed C/E preparation and observable, establish its service budget independently
and collect complete records for the declared rejection protocol.
The construction paper gives the detailed premise map, including the retained negative controls
that bound how the carrier, the response triplets, the feedback class and the kinetic term may be
attached [ 66].
6 Closure: a universe that reads itself into existence
The previous sections built observers, their agreement and the structures they can read oﬀ, and
they stated the executable speciﬁcation as a candidate with a scoreboard. This section returns to
the proposal stated in the introduction and develops it in full. It explains why the speciﬁcation is
exempt from a-priori justiﬁcation and why the theory has no ﬁrst premise, although its exposition
has an entry point.
6.1 The closure hypothesis
The universe is modeled as a timeless, closed, self-referential structure in which the description and
the described system are one object. Let U be a complete inhabited history, let D(U) be an account
of its architecture recoverable through completed records within it, and let R be an admissible
realization of that account constructed inside U. The closure condition is
R
(
D(U)
) ∼= U, (12)
where ∼= is agreement in the observable and interventional structure: the records used to recover
the description, and the responses of the realization to every admitted intervention, ﬁt the world
from which the account was recovered [ 99]. Written as a ﬁxed point of a universe-level closure
operator, the condition reads T (U) = U. Observers infer the speciﬁcation of their world, build
hardware within it, and run that hardware; its execution realizes the same history, including its
builders and their hardware. The identiﬁcation of the construction-side and inhabited-side readings
is the hypothesis. In this sense the structure reads itself into existence.
30

## Page 31

No external computer runs the history and no additional time exists in which the whole is played.
Repair order is a partial order on commits, an observer history is a chain of completed records, and
that order is invariant under strictly increasing regradings, so aﬃne time and proper time require
their own calibration. Fixedness of the complete history permits changing states within it: the
equation concerns the whole history and identiﬁes no last instantaneous state with the universe.
Self-reference earns content only when the return condition restricts the possibilities. On the
real line x ↦→ x + 1 has no ﬁxed point, x ↦→ x ﬁxes every point and x ↦→ (x + 2)/ 2 ﬁxes exactly one.
A candidate therefore has to specify the history domain and the action of T ; existence, uniqueness
and the identiﬁcation of our universe with a ﬁxed point are questions the equation leaves to the
programme. Two ﬁnite results give the condition exact content on declared branches. The packet-
quotient closure theorem proves that on the declared ﬁnite consensus branch the closure map is an
aﬃne idempotent whose ﬁxed points are exactly the packets supported on consensus normal forms
[6]. The universe-level record-closure equation N = log M0(UN ), with UN the regulated observer
system at capacity budget N and M0 the correctable-code capacity of its terminal public record,
is posed exactly, and the all-rung arithmetic proves that the bounded completion class deﬁned by
base agreement, positivity and the carrier bound does not identify it [ 8].
The executable observer gives the condition its ﬁnite meaning. A realization diﬀers from a
description because it carries the described dependencies under intervention: for a described update
τa, realized operation Fa and ﬁxed readout p, the one-step condition p(Fa(s)) = τa(p(s)) extends
by induction to every ﬁnite admitted sequence [ 99]. A playback can match every frame of one
history and fail this test as soon as a retained state is changed. The controls of Section 2.4 perform
exactly this test on the running observer: disconnecting the detector or shuﬄing its bit with the
same frequencies removes the feedback response, while the record-controlled execution keeps it.
6.2 No outside
A theory of an external world must justify its postulates by an argument given in advance or by ﬁt
to data gathered from outside the theory. A self-contained structure oﬀers no vantage point from
which a premise could be imposed. This changes what a premise has to do, and it gives one rule
with three applications.
Justiﬁcation. A premise is justiﬁed when observers recover it and the speciﬁcation that imple-
ments it runs and reproduces their measurements. The carrier, the observer algebra, the repair
law and the closure coordinates are found by observers inside the structure and built by them;
asking for an additional a-priori reason would ask for a view from outside, which the hypothesis
denies. Construction choices therefore need no uniqueness theorem. What they owe is execution
and agreement with what observers measure, which is the content of Section 5.
State selection. The rule also forbids assuming what observers have not recovered. After every
observer-visible constraint has been imposed, the selected state may carry no further information
relative to the recovered reference, so it minimizes the weighted relative entropy ∑
p wpD(ρp∥τp)
over the compatible local state families of the patches p. The theorem library calls this premise
A3. Its ﬁnite objective has exact consequences: every minimizing local density has the union of the
feasible supports, and a feasible witness with failure probability b > 0 gives an explicit ﬂoor on the
selected probability [ 9].
T ransformations. Every change of description that observers can make must be implemented
by the structure’s own operations, since nothing else could implement it. On a carrier this means
31

## Page 32

that each proper recharting is realized from inside the complete reversible response. The library
calls this the endogeneity clause of A2, and it carries real weight: Theorem 4.2 uses it to exclude
the only other twelve-dimensional compact type and to force u(1) ⊕ su(2) ⊕ su(3).
6.3 Agreement and the meaning of “our”
The rule says what needs no outside justiﬁcation. Agreement says what counts as shared. Observers
assign the same operational meaning to accepted data on shared boundaries, naturally under every
restriction, recharting and reﬁnement, and a record is public when it survives every authorized com-
parison (Theorem 3.4). “Our universe” is the world that this public record describes. Agreement
in this sense is a deﬁnition and an acceptance criterion: the shared-meaning row of Table 3 requires
interpretation maps ﬁxed independently of validation outcomes and preserving the declared joint
and conditional statistics, including quantum correlations. It does not require a noncontextual
table of outcomes for incompatible unperformed measurements. The theorem library calls the in-
terpretation clause of this condition A2. The basis reference states the hypothesis, the rule, the
condition and the speciﬁcation formally, with the library’s premise statements and their dictionary
[100].
6.4 An entry point without a starting point
The exposition of this paper enters the loop at the executable speciﬁcation. That choice reﬂects
where observers stand, holding records and inferring, and it makes the account concrete. The
order of justiﬁcation runs the other way: the hypothesis comes ﬁrst, since it is what exempts the
speciﬁcation from outside justiﬁcation. The structure itself has no ﬁrst element in either sense. A
pair of mutually determining equations such as x = (y + 2)/ 2, y = (x + 2)/ 2 ﬁxes x = y = 2 with no
ﬁrst value from which the other is manufactured; a calculator may approach the pair by iteration
and someone may ﬁnd it on a particular day, and neither belongs to the reason the answer is two
[99].
No temporal paradox arises. The closure equation relates parts of one timeless object, with no
global time coordinate along which information travels backward. Observers recover the speciﬁ-
cation and later construct it in their own record order, and both stages belong to the same ﬁxed
point. Repair descent is a well-founded order on authorized updates used to prove termination and
compare schedules; it is neither an observer’s experienced time nor a universe-level clock.
What is found and what is made become two roles in one structure. A rule invented for a game
has consequences its inventor discovers. Here the architecture that observers discover constrains
what they can build, and what they build belongs to the structure from which the discovery is made.
The same reading answers the question why anything exists at the level the hypothesis can answer
it: the question includes its questioner, and a complete inhabited structure whose dependencies
constitute its operation needs no further act of actualization [ 99]. A demand for something in
addition to the complete relational structure is, on this reading, misplaced. If several closures
qualify, the hypothesis admits several worlds, and self-consistency alone has not selected ours.
6.5 What closure ﬁxes, and what observation reads oﬀ
The architecture carries two dimensionless closure coordinates: the screen grain P and the record
capacity N . Under the closure hypothesis these coordinates are ﬁxed points of the closure map,
and the loop also licenses reading them oﬀ: observers locate a constant by measurement, declare
a map, and certify a ﬁxed point of that map inside the measured basin. Whether a certiﬁed ﬁxed
point coincides with the measured value is a diagnostic in the sense of Table 1, not a consequence
32

## Page 33

of the hypothesis. The theory stands on its recovered speciﬁcation whether or not a ﬁrst-principles
evaluation of every coordinate is found.
The screen-grain loop
P = φ + √π/A T (P ),
with AT (P ) the inverse coupling a declared map returns at trial grain P and φ = (1 +
√
5)/ 2,
has one interval-certiﬁed ﬁxed point per declared map. The bare source map settles at α−1 =
136. 994835177413, a relative distance of 3. 0×10−4 from the 2022 CODATA value 137. 035999177(21)
[65]; the self-consistent map, which recomputes the ﬁnite-screen uniﬁed gauge width inside the loop,
settles at P = 1 . 6309682414 and α−1 = 137 . 035660136947, a relative distance of 2. 5 × 10−6, the
gauge width supplying 99.2 percent of the gap. The residual is attributed to same-scheme low-
energy transport to the Thomson endpoint, which neither map computes and which this paper
does not supply [ 10]. Reading the capacity as de Sitter horizon entropy in nats gives
N = 3π
Λ ℓ2
P
, (13)
with ℓ2
P = ~G/c 3, and the capacity branch gives N = 3. 2921 × 10122 and 3. 3001 × 10122 against the
coordinate 3. 3129 × 10122 read back from the weighted Planck base- Λ CDM chain; in the constant
they read Λ = 1 . 0959×10−52 and 1. 0933×10−52 m−2 against 3Ω Λ H 2
0 /c 2 = 1. 0891×10−52 m−2 at the
Planck base-Λ CDM centrals, distances of +0. 62 and +0. 38 percent. Equation ( 13) is the dictionary
that produces the comparison coordinate, so this restates the capacity diagnostic in the constant,
and it consumes the measured ~, G and c through the calibration map (M2). Neither clears the
source-only selection and same-quantity gates. Both comparisons are diagnostics of declared maps
in the sense of Table 1: the maps were exposed before the comparison, the choice of map is itself
declared, and neither comparison carries predictive weight. The executable detector of Section 2.4
consumes P through the coupling g(P ) =
√
4πα(P ), and N bounds the correctable records of a
declared capacity carrier by log M0 ≤ N .
7 Empirical meaning and falsiﬁability
The custody-bound register ﬁxes target deﬁnitions, exclusion thresholds, precision ﬂoors, exposure
classes and decision rules before eligible comparison [ 3]. Two kinds of test follow from the construc-
tion: a propagation anisotropy ﬁxed by the carrier’s symmetry, and an instrument comparison
ﬁxed by the executable observer. Published records also test precise attachments of native records
retrospectively.
7.1 Propagation anisotropy at angular rank six
Angular rank means spherical-harmonic rank j, not a particle spin. The dimension of the invariant
subspace of the sixty proper icosahedral rotations in angular rank j is
mj = 1
60
[
(2j + 1) + 15 U2j(0) + 20 U2j( 1
2 ) + 12 U2j( φ
2 ) + 12 U2j( φ−1
2 )
]
,
with U the Chebyshev polynomials of the second kind, and the exact table gives m1 = · · · = m5 = 0
and m6 = 1 [49].
Theorem 7.1 (Icosahedral angular-rank universality) . Let the carrier kinetics be any operator
λa(k) = a−2 ∑
d w(d) [1 − cos(a k·d)] with a ﬁnite direction multiset and weights invariant under
33

## Page 34

the proper icosahedral rotation group. Then the dispersion is exactly isotropic through angular rank
ﬁve at every order in a, every directional term below angular rank ten is one multiple of the unique
normalized invariant I6, and whenever the weighted sixth-moment coeﬃcient is nonzero the ﬁrst
directional artifact appears at order a4k6, suppressed by exactly 1/ 16 under one binary reﬁnement
at that order. Rank six is the least symmetry-allowed nonzero anisotropic rank in the class.
Proof. The order-a2m−2k2m angular content is an invariant polynomial of degree 2m with harmonic
components at even ranks j ≤ 2m, and the invariant table forces zero at L = 2 , 4, 8 and a one-
dimensional space at L = 6 . The table, the 62-direction stationary census, the orbit veriﬁcation
and the nonvanishing sixth-moment coeﬃcients are certiﬁed in exact arithmetic [ 1].
Proposition 7.2 (Primitive twelve-port branch prediction) . Suppose the intrinsic spatial kinetic
symbol lies in the cosine class of Theorem 7.1 with scale-independent coeﬃcients, the complete
orbit of the twelve port directions ui is its sole hop support through order k6, the carrier scale a is
ﬁnite and positive with the quadratic term normalized to k2, one carrier rest frame is transported
coherently over the experiment, and the tested scalar or polarization-independent sector realizes the
operator. Then
ω2 = 1
2a2
12∑
i=1
[
1 − cos(ak ui ·ˆk)
]
= k2 − a2
20 k4 + a4
840 k6 + 2a4
7875 k6I6(ˆk) + O(a6k8), (14)
with I6(n) = 25
132
∑
i P6(ui ·n), so in a transported frame the branch predicts the scale-free relations
B6
C2
4
= 32
315 , B0
C2
4
= 10
21 , B6
B0
= 16
75 , (15)
and every intrinsic anisotropic coeﬃcient at 1 ≤ j ≤ 5 vanishes.
Proof. The exact carrier moments ∑
i(ui · n)2 = 4 , ∑
i(ui · n)4 = 12 / 5 and ∑
i(ui · n)6 = 12 / 7 +
(64/ 175) I6(n) substituted into the cosine series give the expansion; eliminating a gives the ratios.
A second frozen ray follows from the source seams.
Proposition 7.3 (Source-seam edge ray and conditional propagation branch) . The boundary cur-
rent of the complete thirty-seam source packet maps through the antipodal-odd load readout onto the
even-sum module D6 of Theorem 4.1, with Smith invariants (1, 1, 1, 1, 1, 2), and every directed seam
deﬁnes one translation of the internal record carrier whose complete normalized direction multiset
is the thirty-direction edge orbit. If the feasible move laws and their objective are natural under
the proper carrier action and A3 supplies a unique normalized minimizer, transitivity forces weight
1/ 60 on every directed seam, and the selected internal operator is the source-counting homogeneous
convolution.
Suppose these record translations are physical displacements under one homogeneous position
action, that the complete edge orbit is the sole direct support through the displayed order, that
proper-carrier covariance ﬁxes one common weight, that the same operator acts on one scalar or
polarization-independent physical sector with quadratic term normalized to k2, and that the action
carries the stated gluing, scale, frame, readout, nuisance, and exclusivity data. Its spatial kinetic
symbol is then
Λ a(k, ˆk) = 1
5a2
30∑
j=1
[
1 − cos
(
ak wj ·ˆk
)]
, (16)
34

## Page 35

with C4 = −a2/ 20, B0 = a4/ 840 and B6 = −a4/ 12600, hence B0/C 2
4 = 10/ 21, B6/C 2
4 = −2/ 63 and
B6/B 0 = −1/ 15. The intrinsic anisotropic coeﬃcients at angular ranks one through ﬁve vanish,
and the rank-six vector lies on the rotated I6 orbit.
Proof. Exact integer incidence gives the displayed image and Smith invariants, and the response-
selected Gram pullback gives the completion of Theorem 4.1. The sixty directed labels map two-
to-one onto the thirty signed edge directions, whose unit-normalized moments are ∑
j(wj· n)2 = 10,∑
j(wj· n)4 = 6 and ∑
j(wj· n)6 = 30/ 7 − (2/ 7) I6(n); substitution in ( 16) gives the coeﬃcients and
ratios. The incidence, quotient, orbit binding and moment identities are machine-checked in exact
arithmetic [ 1].
The proposition contains an exact ﬁnite source theorem and a conditional physical branch. The
naturality and minimizer clause supplies a homogeneous internal action whose generator satisﬁes
the positive maximum principle and an exact Dirichlet identity, and whose plane-wave spectrum
is the symbol above. The declared symbol also carries an exact remainder certiﬁcate with no
physical premise: with q = ak, | ˆΛ − P6| ≤ 7
388800 q8 and 19
20 q2 ≤ ˆΛ ≤ q2 uniformly on 0 ≤ q ≤ 1,
where P6 = q2 − q4/ 20 + (1 / 840 − I 6/ 12600)q6; the certiﬁcate turns a into no SI length. The
edge branch and the primitive-port branch have opposite rank-six signs, and their coeﬃcient rays,
custody packages, exposure exclusions and comparison budgets are separate.
Theorem 7.4 (Carrier-class dispersion band) . For the declared positive-weight scalar cosine class
over orbits Os of the sixty proper carrier rotations with radii rs > 0, weights ws > 0 and the
k2 coeﬃcient normalized to one, write Λ norm = k2 + C4k4 + (B0 + B6I6(ˆk))k6 + O(k8) and µm =∑
s ws|Os|rm
s . Then C4 = −(a2/ 20)µ4/µ 2 < 0 for every member; B0/C 2
4 = (10/ 21)µ2µ6/µ 2
4 ≥ 10/ 21
with equality exactly on single-radius members; the anisotropic ranks one through ﬁve vanish; and
B6/B 0 lies in [−16/ 135, 16/ 75], the face and vertex orbits attaining the endpoints and the edge orbit
sitting at −1/ 15.
Proof. Summing over the sixty rotations gives ∑
g((gu)· ˆk)6 = 60
7 + 64
35 I6(ˆu)I6(ˆk) on unit vectors.
The ﬂoor is the Lagrange identity µ2µ6 − µ2
4 = ∑
i<j WiWjr2
i r2
j (r2
i − r2
j )2 ≥ 0 with Ws = ws|Os|,
and the band holds because the rank-six numerator is a positive-weighted mean of seed values of
I6 in [−5/ 9, 1] [1].
A resolved intrinsic dispersion with B0/C 2
4 below 10/ 21, or B6/B 0 outside [−16/ 135, 16/ 75],
or a residue at ranks one through ﬁve, excludes every member of the declared class at once. The
primitive-port relation is a frozen prospective conditional physical-branch prediction: its target
statement, coeﬃcient snapshot, registration record, public commit custody and detached calendar
attestations ﬁx the prospective content, the dated template search whose icosahedral templates re-
turned a family-wide p-value of 0. 64 is ineligible, and a verdict requires a dataset-speciﬁc registration
with a joint likelihood, frames, orientation proﬁle, nuisance model and calibrated threshold. The
branch fails at ﬁve standard deviations or more if an isolated intrinsic C4 is positive, a lower-rank
coeﬃcient is nonzero, or the calibrated proﬁle likelihood excludes the complete manifold C4 < 0,
B0 = 10
21 C2
4 , B6 = 32
315 C2
4 . A null is inconclusive, since the branch supplies no positive lower bound
on a, and a failure rejects the primitive twelve-port propagation branch; it reaches the framework
as a whole only after a stronger source law makes that branch forced and exclusive [ 3].
7.2 An instrument test from the executable observer
Under a common current-record interface every admissible record law has the classical statistics
of C and E. The multiplier g, the coherence between an empty seam and a waiting single load,
35

## Page 36

separates C, E and the law that records only the current. On a ﬁxed coherent two-state source
preparation the readout probability is 3/ 8 + 1
2 Re(g e−iθ): reversing the phase moves C from 5/ 8 to
1/ 8, E predicts 3/ 8 for both, and the law that records only the current predicts 3/ 8 ±
√
2/ 4 [68].
Four ﬁxed settings include common-phase and incoherent controls. The complete-count evaluator
uses 8192 trials per setting, an independently established probability-error budget of 1/ 64 and an
exact sampling certiﬁcate below 1/ 1000; it rejects malformed or incomplete records and reports
no data when no independent counts are supplied. A source-executed implementation runs both
complete 78-state instruments, the four preparations and a complete Lüders readout from twelve-
port control words, retaining every record and attempted trial [ 66]. A sequential design on one
continuing carrier needs neither exact resets nor independent trials: 379 discarded native attempts
before each of 3, 927 trials replace the reset, failed heralds count as trials, and conditional-law
bounds on the dependent transcript give a wrong-candidate probability of at most 3/ 400 under
the exact-source model [ 87]. Synthetic controls exercise the decision rules and establish no physical
verdict; physical encoding, independent calibration and comparison data are the open requirements.
7.3 Native records against public data
Published records can test precise attachments of native records without a new laboratory real-
ization; each test below is retrospective, states its readout hypotheses, and assigns no prospective
verdict.
The diagonal native branch is a classical stochastic process, so under independently chosen
settings and local factorized responses it obeys the CHSH bound S ≤ 2. The deposited records of
the loophole-free Hensen experiment [ 90] reproduce 245 eligible trials, 196 wins and S = 2 . 42250,
and the published ﬁxed-count analysis gives a tail bound near 0. 03917 with its setting-calibration
allowance [ 88]. This rejects the local classical reading of the diagonal branch under that experi-
mental contract. Committed native records form a classical process: the native grammar keeps
diagonal inputs diagonal under every classical feedback history, and two implementations with the
same committed branches, which give 3/ 4 and 1/ 4 under a read before commitment, give 1/ 2 for
both once the record is committed [ 98]. Correlations beyond S = 2 therefore need uncommitted
coherence from declared preparation services, and every admissible record law attenuates it [ 68].
The replay leaves the completed record surface untouched, on which a declared entangled state
reaches the Tsirelson bound of Theorem 3.5, and it does not distinguish C from E.
Event clocks need no external validation, but identifying their readings with a published mea-
surement is a testable hypothesis. Let two native ports, or two distinct regions of one carrier, count
incident repairs, and attach the published 300-second blocks of the Ar 13+/Yb+ optical frequency
ratios of King et al. [ 91] to the normalized count ratio through a gain of at least one, a declared
bracket on the Yb + frequency and an absolute error of 10−12 per block, with direct phase-count
readback. The complete count law, after charging every choice of 2021 subwindow, gain and re-
tained blocks, bounds the probability of the observed block ranges below 10−104 for two native ports
and below 10−22 for every pair of the 3, 433 distinct regional event sets [ 89]. The direct attachment
is rejected together with its readout assumptions; the experiment’s interrogations and servos lie
outside that hypothesis, and the native event clocks are unaﬀected as clocks of the model.
Published trapped-ion records test a supplied experimental model executed through the source
interface. Two calibrated preparation parameters predict 100 reserved complex readouts with per-
spin-mean root-mean-square error 0. 06647, and a changed-potential transfer predicts 125 derived
observations with root-mean-square error 0. 2876 in position-estimator units [ 67]. These compar-
isons test the supplied oscillator model and derive neither its action nor its apparatus from native
repair.
36

## Page 37

7.4 Diagnostics and what fails together
The other quantitative comparisons carry their own decision shapes and do not count as predictions.
The closure roots of Section 6.5 are diagnostics of declared maps. The registered conditional
primordial branch ns = 1 −P/ 48 = 0 . 96602, at the calibrated grain P , lies at the lower 0. 612 percent
position of the oﬃcial ACT DR6, Planck and DESI joint posterior chains, whose mean is 0. 9743 ±
0. 0034 [92, 89]. Those data precede the registration, so this is a retrospective tension under baseline
Λ CDM, not a likelihood ratio, and the registered comparison with a later dataset is unchanged.
A separate ACT DR6 likelihood slice on the measured bandpowers, with the background and
absolute calibration held ﬁxed, prefers a free tilt, ﬁtted at 0. 9773, by ∆ χ2 = 4 . 07; no signiﬁcance
or framework-wide verdict is assigned to it [ 89]. A charged-lepton world average more than three
standard uncertainties from the balanced-circulant center refutes the equal-rank-two tracial reading
of the face-circulant sector; a resolved deep-regime galaxy sample violating degree-one radial scale
covariance or quadrature composition removes the uniqueness of the deep law; and exclusion of
(w0, w a) = ( −1, 0) fails the ﬁxed-capacity branch. The conditional implications share premises,
so they fail in groups: failure of (M4) removes the eﬀective-spacetime and Einstein conclusions
together, failure of (M3) removes the gauge and matter interpretations together, and failure of
(M2) removes every statement in laboratory units. The ﬁnite normal-form, invariant-theory and
algebraic results retain their stated domains under all three.
8 Machine veriﬁcation and reproducibility
Each result maps to an artifact class: a Lean library with no admitted propositions in the de-
pendency closure reported here, including formal premise boundaries and countermodels; exact-
arithmetic code receipts, integer and rational computations and outward-rounded interval certiﬁ-
cates, each emitting a canonical hashed payload; executable observer receipts that replay every
retained result and check its artifact hashes; simulation receipt bundles pinning the source revision,
conﬁguration, seed where stochastic, tolerances and output hashes; veriﬁers that recompute ver-
dicts from clause vectors and reject receipts whose stored verdict disagrees; and adversarial negative
controls, so that every certiﬁcate ships with mutations that must be detected. Machine-readable
records classify every quantitative row as a source result, reconstruction implication, diagnostic,
prospectively ﬁxed test or rejected candidate, and reject any stronger classiﬁcation unsupported by
the recorded dependencies.
Table 4 names the artifact family behind each group of results. The repository, receipts, schemas
and rebuild instructions are public [ 1, 2, 67]. The formal and executable stack checks the claims
assigned to it and enforces their stated boundaries; analytic arguments in this manuscript retain
their displayed proofs, and no part of the stack establishes physical realization.
Table 4: Principal veriﬁcation artifacts and their scope.
Results Principal artifacts
Executable observer, detectors,
interacting process and C/E
instruments
The observer receipts with exhaustive path replays, seeded
executions, feedback controls, exact and interval certiﬁcates, and a
single veriﬁcation entry point that replays every receipt and checks
its hashes [ 67, 66]
37

## Page 38

Table 4 continued
Results Principal artifacts
Overlapping observers, local
clocks, screens and the
continuum operator
The model packages for the observer correspondence, the compiled
clock history, joint and active clocks, repair identiﬁcation, federated
screens and the coherent screen observer, each with exact or interval
receipts, independent oracles and replay; the analytic limit
arguments with their constants and order of limits [ 76, 83]
Public-data comparisons Hash-pinned public records (Bell trials, optical ratio traces, CMB
posterior chains) with exact rational reducers, their stated readout
hypotheses and replay checks [ 88, 89]
Normal forms, stability,
reﬁnement, complexity and the
transactional diamond
The observable-normal-forms library and component-paper
reductions; the local-diamond and Newman arguments of this
manuscript; the ﬁnite transaction-model veriﬁer
Born weights, Lüders selection,
Tsirelson saturation, no-cloning
The event-algebra modules for the interface, conditioning, the bound
and the public record algebra, with per-module axiom audits
Rank-three completion The port-Gram repair band, covariance and frame-quotient modules
with a source-pinned receipt and independent veriﬁer
Gauge Lie type, matrix witness,
matter image and exterior
selection
The screen modules for the carrier action, commutant and incidence
response; the exact-arithmetic port current certiﬁcate; the
trace-balanced kernel and exterior scan modules
Causal precursor, count limit
and proper-time readout
The provenance and geometry modules; the source-record family
receipts with exact all-pairs counts through q = 89 and the sampled
q = 144 level, the carrier realization receipts and event logs, and the
public archive
Null tomography, small-ball
arithmetic and tensor
completion
The Einstein-branch tensor module for null directions, design,
decoder, injectivity and metric ambiguity; explicit null-balance,
Ward, Bianchi and scale premises
Angular-rank theorem, frozen
branches and the dispersion
band
The primitive-port prediction and carrier-class band modules with
mutation tests; the custody-bound prediction register
Exact federation and the
closure loop
The level-six conﬁguration, seeds, seam arrays, loads, per-schedule
ledgers, quotient hashes, standalone veriﬁer and public archive; the
level-ten settlement receipts; the event logs and recovered
speciﬁcations of the canonical and control sources
Dark sector, vacuum coordinate
and closure roots
The dark-sector, deep-proﬁle, per-cut and ﬁxed-capacity modules
with the matched seen-data diagnostic; the capacity closure and
nonidentiﬁability modules; the interval-certiﬁed grain root
9 Discussion and comparison
Operational reconstructions derive quantum structure from information and composition principles
[17, 18, 19]. OPH makes a narrower statement at the quantum stage: it takes the ﬁnite algebra-
state representation as an input and contributes the completed-record surface on which the event,
conditioning and correlation identities operate, together with the consensus theorem that deﬁnes
completion. Quantum Darwinism locates classical objectivity in redundant environmental records
[44], and the completed-consensus record plays the same public role with redundancy replaced by
protected overlap agreement and an explicit repair dynamics. The cellular-automaton programme
shares the ﬁnite substrate [ 45], while OPH locates the quantum identities on the completed record
surface. Wigner’s-friend extensions constrain agents reasoning about one another’s unﬁnished mea-
surements [46], whose premises concern pre-consensus perspectives that the completed surface does
not represent. Relational quantum mechanics treats physical facts as relational [ 42] and the thermal-
38

## Page 39

time programme reads time from modular structure [ 43]; OPH attaches public facts to protected
overlap records, and its executable observer reads modular ﬂow from the actual observation state
of a running process and lets native incident events drive a modular step whose generator is the
native transfer, with a faithful thermal reference derived from closed native histories; passive repair
alone implements no nontrivial modular automorphism. The stationary native state of a ﬁxed-
number sector is tracial, so its modular ﬂow is trivial, while conditioning on a native current event
gives a Palm reference whose coherent modes carry modular frequency log(5/ 4): modular time in
this construction is relative to an event [ 67]. The local-diamond proof sits beside asynchronous
agreement [ 38, 39], and Dijkstra’s self-stabilization separates recovery from continued legitimate
operation [ 40].
The gravitational branch belongs to the thermodynamic and entanglement-based routes to
geometry [ 53, 54, 57], with Bisognano–Wichmann modular ﬂow and half-sided modular inclusions
as established analytic ingredients [ 50, 51, 52]. The route from a null-projected balance and the
Bianchi identity to the full equation, with Λ as an integration constant, is familiar from trace-free
and unimodular formulations [ 55, 56]; what is added is the ﬁnite nine-direction design with exact
constants and the premise contract (G1)–(G6) stated as one implication. The gauge section uses
Doplicher–Roberts reconstruction for the sector group [ 59, 60] and contributes the forcing of the
local Lie type from complete response and internally implemented rechartings.
The closure hypothesis has precedents of a diﬀerent kind. Wheeler pictured a participatory uni-
verse as a self-excited circuit [ 101]. Tegmark’s mathematical-universe hypothesis identiﬁes physical
existence with mathematical structure [ 102], and Nozick proposed that a fundamental principle
might subsume itself [ 103]. OPH adds two demands: the self-subsumption must be inhabited, car-
ried out through records and agents inside the structure, and it must be testable in part, through
an executable speciﬁcation whose observers can be run and whose instruments make predictions.
Lawvere’s ﬁxed-point theorem shows why unrestricted self-representation fails [ 104]; the closure
condition is restricted to the observable and interventional structure that observers can recover.
A useful comparison of uniﬁcation proposals asks how much physical structure follows from
each declared input and which observations distinguish the resulting models; OPH oﬀers a concrete
case with proofs and executable checks attached to individual implications. Formal veriﬁcation mea-
sures deductive reliability, predictive success requires physical realization and eligible measurements,
and neither theorem counts nor numerical coincidences deﬁne a probability that an architecture
describes nature.
10 Conclusion
OPH proposes that physics is what bounded observers can establish together, and that the universe
is the structure in which such observers recover their own speciﬁcation and realize it. Because
nothing lies outside that structure, the speciﬁcation owes no a-priori justiﬁcation; it has to run and
return the world it was recovered from. The speciﬁcation runs. Its observer repairs, records and
acts on its records, its detector draws its coupling from a closure coordinate, and one interacting
quantum process carries ﬂux, charged matter and a scalar ﬁeld through complete measurements
with a certiﬁed feedback response. Overlapping observers on one carrier read a single complete law
of ﬁelds, local clocks and feedback, and joined carriers on a reﬁning screen carry one continuum
operator for Gaussian ﬁelds, their decay and ﬁnite-speed waves. On a declared radial product over
that screen, observers can describe every ﬁnite menu of their readouts, to any tolerance, by one
smooth static (1 + 3)-dimensional Lorentzian ﬁeld theory.
The ﬁnite theorems turn the speciﬁcation’s readouts into familiar structure: public records
39

## Page 40

independent of the repair schedule, the Born rule and the Tsirelson bound on completed records, a
carrier whose slow response has exactly three directions, the Standard Model gauge Lie algebra with
an anomaly-free generation, a source-derived (1 + 3) causal precursor with proper-time ratios from
record counts, and the Einstein equation from a realized null balance. The prospective content is
the frozen dispersion band of Theorem 7.4 and Proposition 7.2: a resolved intrinsic dispersion with
C4 ≥ 0, with B0/C 2
4 < 10/ 21, with B6/B 0 outside [−16/ 135, 16/ 75] or with a nonzero coeﬃcient
at angular ranks one through ﬁve excludes every member of the declared carrier class at once; and
the C/E instrument test, whose decision rule is ﬁxed before any independent count.
The acceptance criteria state what is missing: a reﬁning family of interacting observer processes
whose statistics converge to a 3 + 1 -dimensional quantum ﬁeld theory with the observed matter
and a quantum reference state, a source-selected radial reading of the two-dimensional screen, and
responsive geometry with calibrated clocks on the same histories. The ﬁve maps of Section 5.2 are
consumed rather than constructed, so these results do not establish that the architecture describes
our universe. They specify what a realization must preserve, and the executable observer is the
construction that has to carry it.
Reproducibility statement
All theorems, certiﬁcates, simulation conﬁgurations, receipts, and the formal library are public in
the project repositories [ 1, 2]. The paper release manifest records the release identiﬁer (r2052) with
PDF hashes and sizes; computational receipt bundles pin the applicable source revision, conﬁgura-
tion, seed and output hashes, with clean-checkout rebuild instructions and negative controls; and
machine-readable claim and prediction provenance accompanies the artifacts [ 3]. This manuscript
synthesizes and strengthens results whose longer proofs also appear in public component preprints
on observable normal forms, ﬁnite event algebras, consensus, Einstein reconstruction, gauge struc-
ture and the executable observer [ 4, 5, 6, 9, 11, 66], a relationship disclosed here so that overlap is
visible.
Statements and Declarations
F unding. The authors declare that no external grants or dedicated third-party research funding
were received for the preparation of this manuscript.
Author contributions. Bernhard Mueller developed the OPH concept and mathematical synthe-
sis, wrote the main manuscript, and coordinated the formal and computational evidence. Mario
Poneder developed the executable observer construction, including the native observer, the inter-
acting source process and their certiﬁcates, and contributed the ﬁnite port-current construction
and veriﬁer, gauge and matter certiﬁcates, and proof review of the edge-entropy and Einstein-
normalization branches. Alexander Osika proposed that the screen pixel ratio and the correctable-
record capacity are derivable rather than ﬁtted, supplied the detuning ﬁxed-point equation that de-
termines the pixel ratio, and originated the echosahedral patch-hardware design on which the screen-
microphysics branch is built, in addition to the physical-realization and hardware boundary, pro-
totype framing, and manuscript review. Kai Xue contributed consistency checks, simulator review,
visualization review, and prototype-engineering review. Ben Cassie contributed implementation-
claim and physical-evidence review. Peter Nguyen contributed proof auditing, branch-boundary
analysis, compact-gauge and global-quotient analysis, and manuscript consistency review. Jinwook
Kim developed the constructions, proofs, and counterexamples for the proﬁle partition and struc-
tural characterizations, exact behavior-cut recovery, exact ﬁnite-state proﬁle transport, adaptive
40

## Page 41

repair stratiﬁcation, the cumulative attempt-capacity classiﬁcation, ﬁxed computation-federation
progress, and the ﬁxed-federation execution classiﬁcation, together with the mechanism-variant
comparison kernel and conditional example and the source-derived phase-eﬀect selection. He con-
structed a provably fair emission-order round-robin scheduler with a linear exact-output attempt
bound, established order-sharp quadratic accepted-step bounds for the actual compiler with the
exponential fanout control, and proved the historical weak-relation fair-stuttering no-go and the
weak-fair delay versus bounded-waste horizon separation. He formalized these results in Lean, con-
nected the repair and execution results to the Tower library, and contributed their manuscript,
claim registry, ledger and build records. David Matscheko contributed proof review of consensus
repair, observable normal forms, modular and Einstein algebra, edge entropy, hypercharge, and
the Z6 quotient. Daniel Ari Friedman contributed independent red-team audits of the papers, the
book, the claim and premise ledgers, the Lean library, continuous integration and the custody
records, which led to scoped headline claims, displayed values matched to their certiﬁed receipts
and repaired falsiﬁcation rules, and the independently reviewed programme-obstruction audit whose
controls bound how the carrier, the response triplets, the feedback class and the kinetic term may
be attached. Jonathan Hill contributed Lean formalization and proof auditing of observable nor-
mal forms, reﬁnement and repair results, complexity classiﬁcations, and artifact coverage. William
T. Glynn contributed simulator development, computational implementation, reproducible-build
infrastructure, and release-validation work in the main OPH research repository. Maarten Antonie
Visser contributed consensus-protocol review, physics feedback on the foundations and emergent-
spacetime framing, and the source-derived causal-poset reading of authenticated provenance that
the event-precedence layer implements. Kale Arnav Anirudha contributed consensus-paper mate-
rial, asynchronous-agreement analysis, repair-map deﬁnitions, and formal review of the protocol
surface. Brieuc de La Fournière contributed ﬁnite Lean constructions on the source-derived causal-
order and carrier-symmetry surfaces, including the cover-relation provenance compiler and the an-
tipodal port-action identiﬁcation, and external comparison controls for the charged-lepton branch.
All authors reviewed and approved the manuscript. All authors agree to be accountable for all
aspects of the work.
Corresponding author. Correspondence should be addressed to Bernhard Mueller.
Email: bernhard@floatingpragma.ai.
Data availability . The ﬁnite witnesses, simulation conﬁgurations, receipts, scientiﬁc claim reg-
istries, and release manifests supporting this study are available in the public OPH and simula-
tion repositories [ 1, 2]. Individual receipts and their bundle documentation provide the source
and output provenance for computational claims. The artifact map identiﬁes the level-six exact-
federation archive and its level-ﬁve to level-ten extension, the closure-loop archive, and the causal-
poset archive.
Code availability . The Lean sources, exact-arithmetic veriﬁers, ﬁnite scans, simulation code,
tests, and rebuild instructions are available in the same public repositories [ 1, 2]. The computational
claims in this manuscript identify their principal ﬁles in Section 8.
Ethics approval and consent to participate. Not applicable. The study involved no human
participants, human data, or animals.
Consent for publication. Not applicable.
Competing interests. The authors participate in OPH-related research, software, simulation,
or prototype-development programs and may receive professional or reputational beneﬁt from this
work. These relationships are disclosed. The authors declare no other competing interests.
41

## Page 42

Credited prior work
The results of this paper build on the following established work. Each entry names an established result as it is used
here and cites its sources.
• Icosahedral harmonics, ﬁnite-diﬀerence dispersion, and nonminimal Standard-Model Extension coeﬃcient
searches [ 112, 114, 110, 111].
• Root lattices, graph incidence, transverse projectors, free-vector oscillators, icosahedral edge harmonics, ﬁnite-
diﬀerence symbols, Lorentz-violating threshold kinematics, and nonminimal Standard-Model Extension coeﬃ-
cient searches [ 106, 108, 115, 113, 109, 110, 111].
• Spherical-design moment theory, lattice dispersion expansions, and Lorentz-violation coeﬃcient parametriza-
tions [ 107, 113, 110, 105].
References
[1] FloatingPragma (2026). Observer Patch Holography: Lean library, exact code receipts, ﬁnite certiﬁcates, and
scientiﬁc claim registries. https://github.com/FloatingPragma/observer-patch-holography
[2] Mueller, B. (2026). OPH physics simulator and reproduction source. https://github.com/muellerberndt
/oph- physics- sim . Exact-federation archive: https://github.com/FloatingPragma/observer- patch
- holography/tree/r2052/evidence/exact_federation_L6_canonical_20260909 . Closure-loop archive:
https://github.com/FloatingPragma/observer-patch-holography/tree/r2052/evidence/closure_loop .
Causal-poset archive: https://github.com/FloatingPragma/observer-patch-holography/tree/r2052/evide
nce/source_net_causal_poset. Active reﬁnement-readout revision: https://github.com/muellerberndt/oph
-physics-sim/tree/036b12b608a9ca4c19d69f5e951b3b83a7a51cd8 .
[3] FloatingPragma (2026). Precommitted comparison certiﬁcates, custody records, and the frozen prediction reg-
ister. Immutable source receipt: https://github.com/FloatingPragma/observer- patch- holograph
y/blob/66176656dc1143f9ec50ba1a6e409c403545857f/code/a5_fingerprint/runtime/spin_six
_primitive_port_prediction_receipt.json . Append-only custody packet: https://github.com/F
loatingPragma/observer- patch- holography/tree/bb73f80d459113ccc82d5a5b5219d1dda78e9a7
9/evidence/custody/falsification/frozen_targets/fz11_2026- 07- 31 . Source-seam edge receipt:
https://github.com/FloatingPragma/observer-patch-holography/blob/r2052/code/a5_fingerprint
/runtime/seam_current_edge_prediction_receipt.json. Source-seam custody and decision-rule clariﬁcation:
https://github.com/FloatingPragma/observer-patch-holography/tree/bb73f80d459113ccc82d5a5b5
219d1dda78e9a79/evidence/custody/falsification/frozen_targets/fz12_2026-08-02 . Koide custody
packet: https://github.com/FloatingPragma/observer-patch-holography/tree/bb73f80d459113ccc82d5
a5b5219d1dda78e9a79/evidence/custody/falsification/frozen_targets/fz10_2026-07-28 . Live register:
https://github.com/FloatingPragma/observer-patch-holography/blob/r2052/claims/frozen_predictio
n_register.json.
[4] Mueller, B., Kim, J., Matscheko, D., and Hill, J. (2026). Observation-Determined Normal Forms: Stability,
Obstructions, and Reﬁnement in Constraint and Rewrite Systems . Public manuscript in the same repository
revision as this paper. https://github.com/FloatingPragma/observer-patch-holography/blob/r2052/ext
ra/observable_normal_forms.pdf
[5] Mueller, B. (2026). Veriﬁed Projection-Event Calculus in Lean 4: Bundled Arbitrary-Partition Pinching, Lüders
Retractions, and CHSH Interoperability . Public manuscript and formal artifact. https://github.com/Floatin
gPragma/observer-patch-holography/blob/r2052/extra/machine_checked_finite_event_algebras.pdf
[6] Mueller, B., Xue, K., Kim, J., Anirudha, K. A., Matscheko, D., and Hill, J. (2026). Reality as a Consensus
Protocol: The Fixed-Point Computation That Implements Physics . Public manuscript. https://github.com/F
loatingPragma/observer-patch-holography/blob/r2052/paper/reality_as_consensus_protocol.pdf
[7] Mueller, B., Osika, A., Xue, K., Cassie, B., de La Fournière, B., and Friedman, D. A. (2026). Federated Echosa-
hedral Screen Microphysics: Patch Hardware, Records, and Observer Synchronization in Observer-Patch Holog-
raphy. Public manuscript. https://github.com/FloatingPragma/observer-patch-holography/blob/r2052/p
aper/screen_microphysics_and_observer_synchronization.pdf
[8] Mueller, B., Poneder, M., Osika, A., Xue, K., Cassie, B., Nguyen, P., and Friedman, D. A. (2026). Observers
Are All You Need . Public manuscript. https://github.com/FloatingPragma/observer-patch-holography/bl
ob/r2052/paper/observers_are_all_you_need.pdf
[9] Mueller, B., Poneder, M., Osika, A., Xue, K., Nguyen, P., Visser, M. A., Matscheko, D., and de La Fournière,
B. (2026). Recovering Observer Spacetime and Einstein Dynamics from Overlap Consistency . Public manuscript.
42

## Page 43

https://github.com/FloatingPragma/observer-patch-holography/blob/r2052/paper/recovering_observ
er_spacetime_and_einstein_dynamics_from_overlap_consistency.pdf
[10] Mueller, B., Osika, A., and Friedman, D. A. (2026). The Fine-Structure Constant as an OPH Pixel Fixed Point .
Public manuscript. https://github.com/FloatingPragma/observer-patch-holography/blob/r2052/extra/f
ine_structure_constant_derivation.pdf
[11] Mueller, B., Poneder, M., Osika, A., Xue, K., Nguyen, P., Matscheko, D., and Friedman, D. A. (2026). Deriving
Standard Model Gauge Structure from Observer Overlap Consistency . Public manuscript. https://github.com
/FloatingPragma/observer-patch-holography/blob/r2052/paper/deriving_standard_model_gauge_struc
ture_from_observer_overlap_consistency.pdf
[12] Mueller, B. (2026). The Positive-Chamber Koide Identity for Icosahedral Face Circulants . Public manuscript and
formal artifact. https://github.com/FloatingPragma/observer-patch-holography/blob/r2052/extra/koi
de_identity_from_positive_c3_face_circulants.pdf
[13] Mueller, B. (2026). The de Sitter Time-Advance Sign from a Finite Screen with Fixed Capacity . Public
manuscript. https://github.com/FloatingPragma/observer- patch- holography/blob/r2052/extra/d
e_sitter_time_advance_sign_from_fixed_screen_capacity.pdf
[14] Mueller, B. and Matscheko, D. (2026). Observer-Patch Holography and the Dark Sector: Modular Charge, the
Anomalous Collar Source, and the Deep Galaxy Law . Public manuscript. https://github.com/FloatingPragm
a/observer-patch-holography/blob/r2052/cosmology/oph_dark_matter_paper.pdf
[15] Mueller, B. (2026). Observer-Patch Holography Cosmological Vacuum and Structure Formation: Finite Transfer
Conditions, Fluctuation Ensembles, and Proto-Objects . Public manuscript. https://github.com/FloatingPra
gma/observer-patch-holography/blob/r2052/cosmology/oph_cosmological_vacuum_and_structure_forma
tion.pdf
[16] Mueller, B. and Hill, J. (2026). Explaining the Yang–Mills Mass Gap with Observer-Patch Repair Dynamics: A
Support-Visible OPH Route to the Clay Problem: A Conditional Identity . Public manuscript. https://github
.com/FloatingPragma/observer-patch-holography/blob/r2052/extra/yang_mills_gap_clay_problem.pdf
[17] Hardy, L. (2001). “Quantum theory from ﬁve reasonable axioms. ” arXiv:quant-ph/0101012. https://arxiv.or
g/abs/quant-ph/0101012
[18] Chiribella, G., D’Ariano, G. M., and Perinotti, P. (2011). “Informational derivation of quantum theory. ” Physical
Review A 84, 012311. https://doi.org/10.1103/PhysRevA.84.012311
[19] Masanes, L., and Müller, M. P. (2011). “A derivation of quantum theory from physical requirements. ” New
Journal of Physics 13, 063001. https://doi.org/10.1088/1367-2630/13/6/063001
[20] Gleason, A. M. (1957). “Measures on the closed subspaces of a Hilbert space. ” Journal of Mathematics and
Mechanics 6, 885–893.
[21] Busch, P. (2003). “Quantum states and generalized observables: A simple proof of Gleason’s theorem. ” Physical
Review Letters 91, 120403. https://doi.org/10.1103/PhysRevLett.91.120403
[22] Clauser, J. F., Horne, M. A., Shimony, A., and Holt, R. A. (1969). “Proposed experiment to test local hidden-
variable theories. ” Physical Review Letters 23, 880–884. https://doi.org/10.1103/PhysRevLett.23.880
[23] Cirel’son, B. S. (1980). “Quantum generalizations of Bell’s inequality. ” Letters in Mathematical Physics 4, 93–100.
https://doi.org/10.1007/BF00417500
[24] Landau, L. J. (1987). “On the violation of Bell’s inequality in quantum theory. ” Physics Letters A 120, 54–56.
https://doi.org/10.1016/0375-9601(87)90075-2
[25] Wootters, W. K., and Zurek, W. H. (1982). “A single quantum cannot be cloned. ” Nature 299, 802–803. https:
//doi.org/10.1038/299802a0
[26] Dieks, D. (1982). “Communication by EPR devices. ” Physics Letters A 92, 271–272. https://doi.org/10.101
6/0375-9601(82)90084-6
[27] Lüders, G. (1950). “Über die Zustandsänderung durch den Meßprozeß. ” Annalen der Physik 443, 322–328.
https://doi.org/10.1002/andp.19504430510
[28] Bombelli, L., Lee, J., Meyer, D., and Sorkin, R. D. (1987). “Space-time as a causal set. ” Physical Review Letters
59, 521–524. https://doi.org/10.1103/PhysRevLett.59.521
[29] Surya, S. (2019). “The causal set approach to quantum gravity. ” Living Reviews in Relativity 22, 5. https:
//doi.org/10.1007/s41114-019-0023-1
[30] Major, S., Rideout, D., and Surya, S. (2007). “On recovering continuum topology from a causal set. ” Journal of
Mathematical Physics 48, 032501. https://doi.org/10.1063/1.2435599
[31] Myrheim, J. (1978). “Statistical geometry. ” CERN preprint TH-2538.
[32] Meyer, D. A. (1988). “The Dimension of Causal Sets. ” PhD thesis, Massachusetts Institute of Technology.
43

## Page 44

[33] Benincasa, D. M. T., and Dowker, F. (2010). “The scalar curvature of a causal set. ” Physical Review Letters
104, 181301. https://doi.org/10.1103/PhysRevLett.104.181301
[34] Dowker, F., and Glaser, L. (2013). “Causal set d’Alembertians for various dimensions. ” Classical and Quantum
Gravity 30, 195016. https://doi.org/10.1088/0264-9381/30/19/195016
[35] Hawking, S. W., King, A. R., and McCarthy, P. J. (1976). “A new topology for curved space–time which
incorporates the causal, diﬀerential, and conformal structures. ” Journal of Mathematical Physics 17, 174–181.
https://doi.org/10.1063/1.522874
[36] Malament, D. B. (1977). “The class of continuous timelike curves determines the topology of spacetime. ” Journal
of Mathematical Physics 18, 1399–1404. https://doi.org/10.1063/1.523436
[37] Lamport, L. (1978). “Time, clocks, and the ordering of events in a distributed system. ” Communications of the
ACM 21(7), 558–565. https://doi.org/10.1145/359545.359563
[38] Lamport, L., Shostak, R., and Pease, M. (1982). “The Byzantine generals problem. ” ACM Transactions on
Programming Languages and Systems 4(3), 382–401. https://doi.org/10.1145/357172.357176
[39] Fischer, M. J., Lynch, N. A., and Paterson, M. S. (1985). “Impossibility of distributed consensus with one faulty
process. ” Journal of the ACM 32(2), 374–382. https://doi.org/10.1145/3149.214121
[40] Dijkstra, E. W. (1974). “Self-stabilizing systems in spite of distributed control. ” Communications of the ACM
17(11), 643–644. https://doi.org/10.1145/361179.361202
[41] Newman, M. H. A. (1942). “On theories with a combinatorial deﬁnition of equivalence. ” Annals of Mathematics
43(2), 223–243. https://doi.org/10.2307/1968867
[42] Rovelli, C. (1996). “Relational quantum mechanics. ” International Journal of Theoretical Physics 35, 1637–1678.
https://doi.org/10.1007/BF02302261
[43] Connes, A., and Rovelli, C. (1994). “Von Neumann algebra automorphisms and time-thermodynamics relation
in generally covariant quantum theories. ” Classical and Quantum Gravity 11, 2899–2917. https://doi.org/10
.1088/0264-9381/11/12/007
[44] Zurek, W. H. (2003). “Decoherence, einselection, and the quantum origins of the classical. ” Reviews of Modern
Physics 75, 715–775. https://doi.org/10.1103/RevModPhys.75.715
[45] ’t Hooft, G. (2016). The Cellular Automaton Interpretation of Quantum Mechanics . Fundamental Theories of
Physics 185. Springer, Cham. https://doi.org/10.1007/978-3-319-41285-6
[46] Frauchiger, D., and Renner, R. (2018). “Quantum theory cannot consistently describe the use of itself. ” Nature
Communications 9, 3711. https://doi.org/10.1038/s41467-018-05739-8
[47] Ratcliﬀe, J. G. (2019). Foundations of Hyperbolic Manifolds , 3rd ed. Graduate Texts in Mathematics 149.
Springer, Cham. https://doi.org/10.1007/978-3-030-31597-9
[48] Serre, J.-P. (1977). Linear Representations of Finite Groups . Graduate Texts in Mathematics 42. Springer, New
York. https://doi.org/10.1007/978-1-4684-9458-7
[49] Cohan, N. V. (1958). “The spherical harmonics with the symmetry of the icosahedral group. ” Proceedings of the
Cambridge Philosophical Society 54(1), 28–38. https://doi.org/10.1017/S0305004100033156
[50] Bisognano, J. J., and Wichmann, E. H. (1975). “On the duality condition for a Hermitian scalar ﬁeld. ” Journal
of Mathematical Physics 16, 985–1007. https://doi.org/10.1063/1.522605
[51] Bisognano, J. J., and Wichmann, E. H. (1976). “On the duality condition for quantum ﬁelds. ” Journal of
Mathematical Physics 17, 303–321. https://doi.org/10.1063/1.522898
[52] Wiesbrock, H.-W. (1993). “Half-sided modular inclusions of von Neumann algebras. ” Communications in Math-
ematical Physics 157, 83–92. https://doi.org/10.1007/BF02098019
[53] Jacobson, T. (1995). “Thermodynamics of spacetime: The Einstein equation of state. ” Physical Review Letters
75, 1260–1263. https://doi.org/10.1103/PhysRevLett.75.1260
[54] Jacobson, T. (2016). “Entanglement equilibrium and the Einstein equation. ” Physical Review Letters 116, 201101.
https://doi.org/10.1103/PhysRevLett.116.201101
[55] van der Bij, J. J., van Dam, H., and Ng, Y. J. (1982). “The exchange of massless spin-two particles. ” Physica A
116, 307–320. https://doi.org/10.1016/0378-4371(82)90247-3
[56] Ellis, G. F. R., van Elst, H., Murugan, J., and Uzan, J.-P. (2011). “On the trace-free Einstein equations as a
viable alternative to general relativity. ” Classical and Quantum Gravity 28, 225007. https://doi.org/10.108
8/0264-9381/28/22/225007
[57] Van Raamsdonk, M. (2010). “Building up spacetime with quantum entanglement. ” General Relativity and Grav-
itation 42, 2323–2329. https://doi.org/10.1007/s10714-010-1034-0
[58] Hall, B. C. (2015). Lie Groups, Lie Algebras, and Representations: An Elementary Introduction , 2nd ed. Gradu-
ate Texts in Mathematics 222. Springer, Cham. https://doi.org/10.1007/978-3-319-13467-3
44

## Page 45

[59] Doplicher, S., and Roberts, J. E. (1989). “A new duality theory for compact groups. ” Inventiones Mathematicae
98, 157–218. https://doi.org/10.1007/BF01388849
[60] Doplicher, S., and Roberts, J. E. (1990). “Why there is a ﬁeld algebra with a compact gauge group describing
the superselection structure in particle physics. ” Communications in Mathematical Physics 131, 51–107. https:
//doi.org/10.1007/BF02097680
[61] Georgi, H., and Glashow, S. L. (1974). “Unity of all elementary-particle forces. ” Physical Review Letters 32(8),
438–441. https://doi.org/10.1103/PhysRevLett.32.438
[62] Bouchiat, C., Iliopoulos, J., and Meyer, Ph. (1972). “An anomaly-free version of Weinberg’s model. ” Physics
Letters B 38(7), 519–523. https://doi.org/10.1016/0370-2693(72)90532-1
[63] Tong, D. (2017). “Line operators in the Standard Model. ” Journal of High Energy Physics 07, 104. https:
//doi.org/10.1007/JHEP07(2017)104
[64] Witten, E. (1982). “An SU(2) anomaly. ” Physics Letters B 117, 324–328. https://doi.org/10.1016/0370-269
3(82)90728-6
[65] Mohr, P. J., Newell, D. B., Taylor, B. N., and Tiesinga, E. (2025). “CODATA recommended values of the funda-
mental physical constants: 2022. ” Reviews of Modern Physics 97, 025002. https://doi.org/10.1103/RevModPh
ys.97.025002
[66] Mueller, B., Poneder, M., Kim, J., and Friedman, D. A. (2026). An Executable Observer for Observer Patch
Holography: Speciﬁcation, Certiﬁed Components and Acceptance Criteria . Public manuscript. https://github
.com/FloatingPragma/observer-patch-holography/blob/r2052/paper/executable_observer_constructio
n.pdf
[67] FloatingPragma (2026). The runnable OPH model: observer implementation, tests, receipts and construction
guide. https://github.com/FloatingPragma/observer-patch-holography/tree/r2052/model
[68] FloatingPragma (2026). Admissible record laws for native repair. https://github.com/FloatingPragma/obse
rver-patch-holography/blob/r2052/model/record_laws/README.md
[69] FloatingPragma (2026). Native quantum mixing and the full-carrier memory boundary. https://github.com/F
loatingPragma/observer-patch-holography/blob/r2052/model/observer/QUANTUM_MIXING.md
[70] FloatingPragma (2026). What complete native records can recover. https://github.com/FloatingPragma/ob
server-patch-holography/blob/r2052/model/observer/RECORD_RECOVERY.md
[71] FloatingPragma (2026). The inﬁnite-occupation sharp-readout history of the common interacting source. https:
//github.com/FloatingPragma/observer-patch-holography/blob/r2052/model/common_source/INFINITE_
HISTORY.md
[72] FloatingPragma (2026). Complete comparisons with the actual ﬁnite pointer. https://github.com/FloatingP
ragma/observer-patch-holography/blob/r2052/model/common_source/READOUT.md
[73] FloatingPragma (2026). The ﬁnite executable observer speciﬁcation. https://github.com/FloatingPragma/ob
server-patch-holography/blob/r2052/model/common_source/COMPOSITION.md
[74] FloatingPragma (2026). Coherent native repairs around one continuing quantum history. https://github.com
/FloatingPragma/observer-patch-holography/blob/r2052/model/common_source/QUANTUM_NATIVE.md
[75] FloatingPragma (2026). Active encoded memory in the continuing observer. https://github.com/FloatingP
ragma/observer-patch-holography/blob/r2052/model/common_source/ENCODED_MEMORY.md
[76] FloatingPragma (2026). One process, overlapping observer records and an eﬀective transfer. https://github.c
om/FloatingPragma/observer-patch-holography/blob/r2052/model/observer_correspondence/README.md
[77] FloatingPragma (2026). Source-compiled apparatus in a complete observer-clock history. https://github.com
/FloatingPragma/observer-patch-holography/blob/r2052/model/compiled_clock_history/README.md
[78] FloatingPragma (2026). Joint local clocks and quantum ﬁelds with explicit memory. https://github.com/Flo
atingPragma/observer-patch-holography/blob/r2052/model/joint_clock_process/README.md
[79] FloatingPragma (2026). Fixed observer predictions and complete-count comparison: modular clock recovery and
native repair response. https://github.com/FloatingPragma/observer-patch-holography/blob/r2052/mod
el/predictions/README.md
[80] FloatingPragma (2026). Identifying the complete current-resolving repair law. https://github.com/FloatingP
ragma/observer-patch-holography/blob/r2052/model/repair_identification/README.md
[81] FloatingPragma (2026). A local action for native quantum observations. https://github.com/FloatingPragm
a/observer-patch-holography/blob/r2052/model/native_action/README.md
[82] FloatingPragma (2026). Coherent observer outputs on matched source screens. https://github.com/Floatin
gPragma/observer-patch-holography/blob/r2052/model/coherent_screen_observer/README.md
45

## Page 46

[83] FloatingPragma (2026). A spatial ﬂuctuation ﬁeld from one native history: the common spatial generator, its
locality and spectral dimension, and the boundary-trace test. https://github.com/FloatingPragma/observer
-patch-holography/blob/r2052/model/physical_space_time/OCCUPATION_FIELD.md
[84] FloatingPragma (2026). Joint time and ﬁeld laws on matched source screens. https://github.com/FloatingP
ragma/observer-patch-holography/blob/r2052/model/observer_time_fields/REFINEMENT.md
[85] FloatingPragma (2026). Tangent ﬁrst laws and ﬁnite entropy changes: native reference entropy and region
boundaries. https://github.com/FloatingPragma/observer-patch-holography/blob/r2052/code/geometry
/ENTROPY_FIRST_LAW_AUDIT.md
[86] FloatingPragma (2026). Exact native causal cones. https://github.com/FloatingPragma/observer-patch-h
olography/blob/r2052/model/common_source/CAUSALITY.md
[87] FloatingPragma (2026). A ﬁnite sequential native-law prediction protocol. https://github.com/FloatingPra
gma/observer-patch-holography/blob/r2052/model/closed_prediction_protocol/README.md
[88] FloatingPragma (2026). Classical record responses and published Bell measurements. https://github.com/Flo
atingPragma/observer-patch-holography/blob/r2052/model/predictions/BELL_RECORDS.md
[89] FloatingPragma (2026). OPH predictions and public-data comparisons. https://github.com/FloatingPragm
a/observer-patch-holography/blob/r2052/docs/research/PUBLIC_DATA_PREDICTION_AUDIT.md
[90] Hensen, B. et al. (2015). Loophole-free Bell inequality violation using electron spins separated by 1.3 kilometres.
Nature 526, 682–686. https://doi.org/10.1038/nature15759
[91] King, S. A. et al. (2022). An optical atomic clock based on a highly charged ion. Nature 611, 43–47. https:
//doi.org/10.1038/s41586-022-05245-4
[92] Louis, T. et al. (2025). The Atacama Cosmology Telescope: DR6 power spectra, likelihoods and ΛCDM param-
eters. arXiv:2503.14452. https://arxiv.org/abs/2503.14452
[93] Mosco, U. (1994). Composite media and asymptotic Dirichlet forms. J. Funct. Anal. 123, 368–421.
[94] Fukushima, M., Oshima, Y. and Takeda, M. (2011). Dirichlet Forms and Symmetric Markov Processes , 2nd
edition. De Gruyter, Berlin.
[95] Li, X. and Ding, P. (2017). General forms of ﬁnite population central limit theorems with applications to causal
inference. J. Amer. Statist. Assoc. 112, 1759–1769.
[96] Coulhon, T. and Sikora, A. (2008). Gaussian heat kernel upper bounds via the Phragmén–Lindelöf theorem.
Proc. London Math. Soc. 96, 507–544.
[97] Girouard, A. and Polterovich, I. (2017). Spectral geometry of the Steklov problem. J. Spectral Theory 7, 321–359.
[98] FloatingPragma (2026). Selecting one source, interface and repair law . Model document. https://github.com
/FloatingPragma/observer-patch-holography/blob/r2052/model/SOURCE_SELECTION.md
[99] FloatingPragma (2026). Why Does Anything Exist, and Why Is It Like This? OPH essay B. https://github.c
om/FloatingPragma/observer-patch-holography/blob/r2052/essays/B-why-does-anything-exist.pdf
[100] FloatingPragma (2026). The OPH Basis : the closure hypothesis, the no-outside rule, the agreement condition,
the executable speciﬁcation, and the theorem-library premises. https://github.com/FloatingPragma/observ
er-patch-holography/blob/r2052/docs/BASIS.md
[101] Wheeler, J. A. (1983). “Law without law. ” In J. A. Wheeler and W. H. Zurek (eds.), Quantum Theory and
Measurement, 182–213. Princeton University Press.
[102] Tegmark, M. (2008). “The mathematical universe. ” Foundations of Physics 38, 101–150. https://doi.org/10
.1007/s10701-007-9186-9
[103] Nozick, R. (1981). Philosophical Explanations . Harvard University Press.
[104] Lawvere, F. W. (1969). “Diagonal arguments and cartesian closed categories. ” Lecture Notes in Mathematics
92, 134–145. https://doi.org/10.1007/BFb0080769
[105] Colladay, D., and Kostelecký, V. A. (1998). “Lorentz-violating extension of the standard model. ” Physical
Review D 58, 116002. https://doi.org/10.1103/PhysRevD.58.116002
[106] Conway, J. H., and Sloane, N. J. A. (1999). Sphere Packings, Lattices and Groups . Springer New York. https:
//doi.org/10.1007/978-1-4757-6568-7
[107] Delsarte, P., Goethals, J. M., and Seidel, J. J. (1977). “Spherical codes and designs. ” Geometriae Dedicata 6,
363–388. https://doi.org/10.1007/BF03187604
[108] Godsil, C., and Royle, G. (2001). Algebraic Graph Theory . Springer. https://doi.org/10.1007/978-1-461
3-0163-9
[109] Jacobson, T., Liberati, S., and Mattingly, D. (2003). “Threshold eﬀects and Planck scale Lorentz violation:
Combined constraints from high energy astrophysics. ” Physical Review D 67, 124011. https://doi.org/10.110
3/PhysRevD.67.124011
46

## Page 47

[110] Kostelecký, V. A., and Mewes, M. (2009). “Electrodynamics with Lorentz-violating operators of arbitrary
dimension. ” Physical Review D 80, 015020. https://doi.org/10.1103/PhysRevD.80.015020
[111] Kostelecký, V. A., and Russell, N. (2011). “Data tables for Lorentz and CPT violation. ” Reviews of Modern
Physics 83, 11–31. https://doi.org/10.1103/RevModPhys.83.11
[112] Meyer, B. (1954). “On the Symmetries of Spherical Harmonics. ” Canadian Journal of Mathematics 6, 135–157.
https://doi.org/10.4153/CJM-1954-016-2
[113] Strikwerda, J. C. (2004). Finite Diﬀerence Schemes and Partial Diﬀerential Equations . SIAM. https://doi.
org/10.1137/1.9780898717938
[114] Trefethen, L. N. (1982). “Group Velocity in Finite Diﬀerence Schemes. ” SIAM Review 24, 113–136. https:
//doi.org/10.1137/1024038
[115] Weinberg, S. (1995). The Quantum Theory of Fields, Volume 1: Foundations . Cambridge University Press.
https://doi.org/10.1017/CBO9781139644167
AI Assistance Disclosure
This research project used research-grade commercial models, including Anthropic’s Fable and
OpenAI’s GPT-6 Astra, for research support, software development, editing, and synthesis. The
authors are responsible for the paper’s claims, methods, and ﬁnal text.
47


---
*Extraction method: pypdf*