---
source_pdf: Density_Field_Dynamics__v5_0__Extended_Derivations_and_Frontier_Predictions.pdf
title: "Density Field Dynamics v5.0: Extended Derivations and Frontier Predictions"
version: "v5.0"
date: "2026-10-26"
site: https://densityfielddynamics.com/
author: Gary Alcock
framework: Density Field Dynamics (DFD)
doi: "10.5281/zenodo.18066593"
companion_to: papers/Density_Field_Dynamics__A_Complete_Unified_Theory__v5_0.md
format_note: "Markdown extracted mechanically from the source PDF for clean AI ingestion. No editorial changes; mathematical typography in the PDF is authoritative. This is the 77-page companion volume to the v5.0 main volume; the main volume's statement of the theory and its list of principal corrections are authoritative."
---

Density Field Dynamics v5.0:
Extended Derivations and Frontier Predictions
Gary Alcock
Density Field Dynamics Program
(Dated: v5.0 — October 2026)

This is the complete extended-derivation companion to
the main paper Density Field Dynamics: A Complete
Unified Theory. It collects the proofs, multi-route derivations, frontier predictions, and high-precision numerical
results that support the theory but are too lengthy for
the main text: closures and promotions, the new-territory
frontier (particle, astrophysical, and cosmological predictions), the CKM/τ and series IV–VI derivations, the
mathematical-foundation and quantum-gravity closures,
the multi-route fortifications and sharpenings, and the
wave-sector closure (App. WS). It is not the statement
of the theory; the axioms, theorems, predictions, and
falsifiers are in the main paper, whose list of principal corrections is authoritative. The intermediate development
record is deliberately omitted. Appendix letters other
than AH point to the main paper (App. WS is printed in
both volumes); references within this volume are internal.

CONTENTS

AH. Closures and New Territories
4
1. Purpose and Scope
4
2. Closure and Status of Blocking Open
Problems
4
a. Vacuum Energy Feedback Stability
Theorem
4
b. Strong-Field Symmetric-Hyperbolic
Reformulation
4
c. Convergence Kernel Theorem
5
d. Cluster First-Principles Pipeline
5
e. Internal-Manifold Axiom Reduction
6
3. Statements
6
a. Padé Identity All-Orders Theorem and
Firewall
6
b. Microsector Finiteness Theorems
7
c. Higgs Hierarchy
7
d. cT = c√Structural Theorem
8
e. a0 = 2 α cH0 Derivation Chain
8
f. Topological H0 and Cosmological
Constant
9
g. Penrose Superposition Paradox
Dissolution
9
h. Fine-Structure Constant Closed-Form
Identity
9
i. Charged Fermion Mass Formula
Tightening
10
j. Neutrino Sector: Normal Ordering and
δCP = −π/2
10
4. New Theoretical Territories
10

a. Topological Pre-Inflation
10
b. Baryogenesis from S 3 Topology
11
c. The transport axiom and ωb
14
d. Transit Times and the Horizon
15
e. Quantum-Gravity Amplitude Finiteness 15
f. Neutron Stars
16
g. Primordial Gravitational Waves
16
h. Connes Spectral-Triple Connection
16
i. Path Integral Quantization
16
j. Topology Change of Optical Slices
(Optional)
16
5. Frontier Territories: Particle and
Astrophysical Predictions
17
a. Muon and Electron g − 2: Structurally
Negligible
17
b. QCD Confinement and Glueball
Spectrum
17
c. Electroweak Phase Transition and
Sphaleron Rate
18
d. Multimessenger Binary Neutron Star
Signatures
18
e. Proton Lifetime
18
f. Holographic Entropy Bounds
19
g. Ultra-High-Energy Cosmic Rays and
Photon Dispersion
19
6. Standard Model Lagrangian, BSM
Constraints, and Cosmology
19
a. Vacuum Stability
19
b. Topological Pre-Inflation Status
20
c. CKM Per-Channel Agreement
20
d. Full Standard Model Lagrangian from
Microsector
20
e. Beyond-Standard-Model Constraints
20
f. Strong-CP Closure: Tree Level,
Loop-Bounded
22
g. Precision QED Beyond g − 2
22
h. Lepton Flavor Violation: Structurally
Negligible
22
i. Galaxy Formation in DFD: the High-z
Halo Abundance
23
j. Type Ia Supernovae and the Hubble
Diagram Plateau
23
k. Big Bang Nucleosynthesis: Standard
Abundances
23
l. NANOGrav PTA: SMBH-Binary
Spectrum, Cosmic Strings Excluded
23
m. Stochastic ψ-PDE: Pathwise Existence
and Spectral Gap
24
n. Modular Tensor Category: the SU (2)58
Chern–Simons TQFT
24

2
o. Spectral Data of DFD
24
p. Reionization in DFD
24
q. Background–Perturbation Decomposition
and Scale Partition
25
r. ∆ψ Growth Bridge
25
s. Action-Principle GHY-Analog Boundary
Term
26
7. CKM, Higgs Self-Coupling, Halo Profiles,
and Anomaly Cancellation
26
a. CKM ρ̄ Cayley-Graph Half-Integer
Correction (Conjectural Selection-Rule
Extension, Not a Derived Theorem)
26
b. Pion Decay Constant from GMOR
27
c. Higgs Self-Coupling κλ
28
d. Strict Lepton-Flavor Universality and
B-Meson Predictions
28
e. The Baryons’ Phantom Halo Profile: Flat
Cores
29
f. Sunyaev-Zel’dovich Effect and Cluster
Counts
29
g. Cosmic Shear S8
29
h. Galaxy-Galaxy Lensing and Time-Delay
Cosmography
30
i. Anomaly Cancellation via Atiyah-Singer
Index = 60
31
j. Modular Invariance and Photon-Sphere
Casimir Shift
32
k. DFD and a Candidate Topological
Quantum Error-Correcting Code
32
l. BRST Cohomology and Physical Hilbert
Space
32
m. TPI Conjectured-Extension Status
(CMAH Replacement Available)
32
8. Tau Precision, Heavy-Ion QGP, Kaon
Physics, and DESI Dark Energy
33
a. Tau-Lepton Precision Gauntlet
33
b. Heavy-Ion QGP Critical Temperature
and Viscosity
33
c. QCD String Tension and Lattice
Discriminator
34
d. Casimir Scaling at Zero Free Parameters 34
e. Kaon Physics: NA62 and ε’/ε
Inheritance
35
f. DESI Dark Energy: one constant vacuum
term, w = −1
35
g. BBN ηB : Leptogenesis Estimate (Not a
Prediction — ηB is a Measured Input) 35
h. 21-cm HERA/SKA Power-Spectrum
Fingerprint
36
i. White Dwarf Self-Consistency Theorem 36
j. IceCube Neutrino Astronomy
Inheritance
36
k. Geometric Langlands Connection
36
l. Twistor-Theoretic Identification
37
m. McKay Correspondence: 2I ↔ E8
37
n. Results Checked by Independent
Recomputation
37
9. Additional Theorems and Proofs (Series IV) 37

a. Chiral Condensate from Berry-Volume
Spectral Integral
37
b. Black Hole Entropy from DFD Topology 38
c. No-Native-Inflaton Theorem
38
d. SUSY Forbidden in the Microsector
39
e. Complete Naturalness Theorem
40
f. Holographic Principle: Photon-Sphere
Bulk-Boundary Duality
40
g. CKM ρ̄: Both Paths Recorded
41
h. QECC: Explicit Pauli-Stabilizer
Construction
41
i. Microsector Mass Spectrum:
Direct-Detection Locked Out
41
j. Amplitude Bootstrap Saturation
42
k. Dark-Matter Null Portfolio
42
l. Axiomatic Foundation
42
10. Additional Theorems and Proofs (Series V) 43
a. Master DFD Lagrangian:
Single-Equation Statement
43
b. Single Master Topological Invariant
43
c. Wheeler-DeWitt Equation in DFD
43
d. Single-Vacuum Theorem (No
Landscape)
44
e. Vacuum Deposit: All-Orders Theorem 44
f. EW-Planck Hierarchy: No-Tuning
Theorem
45
g. All-Orders QG Amplitude Finiteness
45
h. Quantum Measurement Problem,
Structural Reduction
45
i. Direct Microsector Lab Analog
45
j. AdS/CFT Without AdS: A
Photon-Sphere Holographic Duality
46
k. Two-Axiom Minimal Statement of DFD 46
11. Additional Theorems and Proofs (Series VI) 46
a. Closed Form for the Fine-Structure
Constant
46
b. Sub-meV Neutrino Spectrum Closure
47
c. Comparative Structural Profile of DFD
vs. Four Major UV-Completion
Programs
47
d. Cross-Cutting Consistency Theorem
48
e. Sector-Resolved Casimir Modification
(Leading Order)
48
f. Quantum-Classical Mass Scale
49
g. Mathematical-Structure Connections
(Riemann, Moonshine, Triality)
49
h. Audit of Existing Data
49
12. Mathematical-Foundation Closures and
High-Precision Predictions
49
a. Atiyah–LeBrun–Witten Triality
50
b. Wheeler–DeWitt Closed-Form Ground
State, Derivation Status
50
c. Quantum Measurement Problem,
Structural Reduction
51
d. Single-Vacuum Theorem, Rigidity
Layers
51
e. AdS/CFT-without-AdS Bulk-Boundary
Identity
52

3
f. QECC (partially established):
Pauli-Stabilizer Structure,
Literal-Hamiltonian Realization
Conditional
52
g. Light-Fermion Generation-1 Mass
Refinement
53
h. Sphaleron-Rate Prefactor
53
i. PTA Amplitudes and the
Supermassive-Binary Background
54
j. Structural Negative Results
54
13. Origin, Multi-Route Electromagnetism,
Quantum Gravity, and Materials-Science
Predictions
54
a. Origin of the Universe in DFD
54
b. Four Routes to Ω = 60
55
c. Origin of CP 2 × S 3 : Eight-Layer
Manifold Rigidity
55
d. Master Derivation Tree of the
Standard-Model Constants
56
e. Four Presentations of α−1
56
f. Non-Perturbative Quantum-Gravity Path
Integral
57
g. Observables and the Physical Metric
57
h. Quantum-Gravity S-Matrix
57
i. Three-Layer Back-Reaction System
57
j. Near-Room-Temperature
Superconductivity Prediction
58
k. ψ-Protected Topological Phase (ψPT) 58
l. Metamaterial Analog Program for DFD 59
m. Multiferroic Magnetoelectric Memory
Effect
59
n. DFD Completeness Theorem on 40
Fundamental Questions
59
o. Single Master Equation for Origin and
Evolution
59
14. Translation Dictionary: DFD-Native Objects
vs. Graviton Language
60
15. Closures of Remaining Partial-Grade
Theorems
60
a. Yukawa Precision: Running Does Not
Close the Residuals
60
b. No optical-screen reconciliation of the
Hubble clocks
61
c. Soft-Graviton Sub-Leading Corrections 61
d. CKM ρ̄ Path A Canonical Closure
61
e. Frame-Theorem ksplit Smooth
Interpolation
62
16. Fortifying Theorems for Multi-Route
Derivations
62
a. Five Presentations of α−1
62
b. Keff : Zero in the Microsector
62
c. Completeness Theorem Cardinality
34/1/5
63
d. Bundle (a, n) = (9, 5) from χ = 60: a
Consistency Check
63
e. Priming Family
63
f. Branch B Exponents: A Forcing
Argument
63

g. Five-Presentation Λ ∝ α57 MP4
Convergence
64
h. (3,2,1) Gauge Bundle: Selected, Not
Forced
64
i. Independence Theorem for the α
Presentations
64
j. BMS Strict Identity at All Orders
65
k. Structural-Derivation Note
65
17. Sharpenings and New Predictions
65
a. Categorical Analysis of the α
Presentations
65
b. Solomonoff–MDL Bayesian Model
Comparison
66
c. Burnside Identity Forced at
Representation-Theoretic Level
66
d. Pre-Registration Tier Classification of
Observational Signals
66
e. The Origin Question and the
Wheeler–DeWitt State
67
f. Anthropic-Principle Redundancy
Theorem
67
g. String-Landscape Exclusion Theorem
67
h. QCD Axion Structurally Excluded
67
i. Muon-to-Electron Mass Ratio
68
j. Higher-Form Symmetries of the
Microsector
68
k. The a6 Heat-Kernel Coefficient
68
l. Geometric Langlands Correspondence on
the DFD Slice
68
m. Twistor Encoding (Conditional)
69
n. The Continuum Limit via
Bochner–Minlos
69
o. Two-Loop Graviton-Scattering
Coefficient
70
p. Subleading Corrections to
Trans-Planckian BH Formation
70
WS. Wave-Sector Closure: Elliptic Scalar, Derived
Gravitomagnetism, δrad = 0, and Ringdown
71
1. Axiom E (Elliptic Constraint Reading)
71
2. The verified 1PN source and the
Standard-Gauge Coincidence Lemma
72
3. Vector-Sector Completion: g0i derived, not
matched
72
4. δrad = 0 with a mechanism; scalar memory;
falsifiers
76
5. Ringdown and echoes: the optical and tensor
channels
76
6. Consistency notes on other sections
76
7. What remains open in this sector
77

4
Appendix AH: Closures and New Territories
1.

Purpose and Scope

This appendix collects the extended results. Each subsection presents the main theorem, the central proof, the
principal numerical consequence, and the falsification criterion. Where the proof is constructive, we display the
construction explicitly; where it is analytic, we display the
central inequality. Where an independent recomputation
reached a different conclusion from the derivation, the
result is recorded immediately after the proof in a Status
note, so the reader can weigh the evidence.

2.

Closure and Status of Blocking Open Problems
a.

Vacuum Energy Feedback Stability Theorem

The Sec. IX.H heuristic “λ ∼ 10113 violently unstable” is closed in the dimensionless response variable
(Sec. AH 6 a); the status note below records an independent recomputation.
Theorem AH.1 (Wilsonian Regulation by Microsector
(1)
Finiteness). Let VCW be the one-loop Coleman–Weinberg
correction to the ψ-effective potential, computed against
the dynamical microsector spectrum on CP 2 × S 3 at
Toeplitz cutoff kmax = 60. Then


60
Mn2 (ψ) 3
1 X 4
(1)
M
(ψ)
log
−
,
VCW =
64π 2n=1 n
Λ2top
2
√
where Λtop = α MP is the largest Toeplitz eigenvalue
scale, and the sum is over dim Hmicro = 60 modes.
Proof. The Berezin–Toeplitz quantization of E = O(9) ⊕
O⊕5 over CP 2 at level kmax = 60 gives a Hilbert space
Hmicro = H 0 (CP 2 , E) whose dimension is computed by
Hirzebruch–Riemann–Roch: dim Hmicro = χ(CP 2 , E) =
55 + 5 = 60. This is finite, not regulated infinite. The
one-loop effective potential is the standard functional
determinant


(1)
VCW (ψ) = 21 TrHmicro log D̂2 (ψ)/µ2 ,
where D̂ is the gauge-covariant Dirac operator on E. Because Hmicro is 60-dimensional, the trace is a finite sum
over 60 mode masses Mn2 (ψ). Performing the Schwinger
proper-time integral for each mode against the standard
MS-equivalent scheme (proper-time cutoff Λ2top ) yields
the displayed Coleman–Weinberg formula. Because each
Mn2 (ψ) ≤ Λ2top by construction (largest Toeplitz eigenvalue), every summand is bounded, and the integral is a
finite trace, not a continuum integral to ∞.
Lemma AH.2 (One-Loop Per-Mode α-Cancellation).
Under uniform gauge normalization, each of the 57
nonzero modes contributes one factor of α to the Gaussian determinant ratio, as at tree level (Lemma O.5 of

the main volume); that this survives at one loop is stated,
not proved here.
(1)

Corollary AH.3 (Vacuum Stability). |δVCW /MP4 | ≤
O(α2 ), and the exponent in GℏH02 /c5 = Cα57 (App. O)
survives quantization with corrections bounded by α2 ,
given Lemma AH.2; this does not derive measured C.
(1)

Proof. By Theorem AH.1 the one-loop VCW is bounded
(1)
by 60 · Λ4top · O(1) = 60α2 MP4 , so |δVCW /MP4 | ≤ O(α2 ).
The per-mode count of Lemma AH.2, as stated, preserves
the exponent in GℏH02 /c5 = Cα57 under quantization
with relative correction δ(α57 )/α57 ≤ O(α2 ); C remains
measured.
Remark AH.4 (Status note). An independent recomputation grades this closure questionable. It agrees
that the partition-function ratio on Hmicro is finite and
scheme-independent, but finds that the 4D Coleman–
Weinberg loop integral over external momenta retains
a UV-divergent shell beyond the Toeplitz truncation, giving δm2ψ ∼ α2 MP2 rather than the much smaller bound
claimed. The discrepancy turns on whether the response
variable is ρvac (master-text reading, instability) or δα/α
in dimensionless units (Gordon optical reading, stability
automatic). Sec. AH 6 a displays the dimensionless bound
and traces the 10113 of Sec. IX.H to the use of ρc in its
denominator; both readings are kept there.
In the dimensionless (Gordon optical) reading the “violently unstable” Sec. IX.H warning does not apply: the
feedback is stable (Corollary AH.3).

b.

Strong-Field Symmetric-Hyperbolic Reformulation

Nonlinear strong-field evolution is locally well-posed
wherever ∇ψ ̸= 0, for the exact law as for its approximant
(main volume, well-posedness section), and the linearised
evolution is well-posed on every static background, including points where ∇ψ = 0. What stays open is nonlinear
evolution through points where ∇ψ = 0, where the symmetriser A0 degenerates and both scalar speeds vanish.
The scalar characteristic speeds of the exponential response reach 1.017 c (SOURCE) and 1.066 c (TOTAL)
in the transition regime; every characteristic advances
the absolute time t of the preferred foliation, so causality
holds and only light-cone containment fails (main volume,
well-posedness section).
Theorem AH.5 (15-Component Symmetric-Hyperbolic
TT
System). Let U ≡ (ψ, πψ , ϕi , hTT
ij , kij ) be the 15component evolution vector. The DFD coupled trace+TT
system can be written
A0 (U ) ∂t U + Aj (U ) ∂j U = S(U ),
with explicitly symmetric A0 and symmetric Aj , A0 positive definite wherever ∇ψ ̸= 0. The principal symbol
σ(ξ) = (A0 )−1 Aj ξj has real eigenvalues for all spacelike

5
ξ. Local well-posedness in H s for s > 5/2 follows by Kato–
Friedrichs on any region where A0 is uniformly positive
definite.
Proof. Write the Lagrangian L = Lψ + LTT + Lint . The
trace sector contributes the variables (ψ, πψ , ϕi ) where
πψ = ψ̇ and ϕi = ∂i ψ. The TT sector contributes
TT
TT
TT
(hTT
ij , kij ) with kij = ḣij . The O(3) block-diagonality
of Theorem AH.20 implies no derivative cross-terms between trace and TT, so the Hessian matrix A0 is blockdiagonal, positive definite wherever ϕ ̸= 0 
(since µ(0) = 0),
with the trace block diag 1, c−2 , Da(ϕ) for (ψ, πψ , ϕi ),
where Da is the flux Jacobian with eigenvalues µ (twice)
1
and (xµ)′ , and 32πG
δij δkl for each TT pair. The spacederivative matrices Aj are read off from ∂j in the EL
equations and inherit the same block structure. By construction each Aj is symmetric. Compute the eigenvalues
of σ(ξ) on each block: the trace block gives the scalarsector characteristic speeds analysed in the main volume’s
well-posedness section (real wherever ϕ ̸= 0); TT gives
±c|ξ| (Theorem AH.20; light and gravitational waves
share one metric, App. WS). Kato’s theorem (1975) for
quasilinear symmetric-hyperbolic systems with C 1 coefficients gives local existence in H s for s > d/2 + 1 = 5/2 in
d = 3 space wherever A0 is uniformly positive definite.
Theorem AH.6 (Constraint Propagation). The kinematic trace-sector constraint Cψ ≡ πψ − ψ̇ and the TTTT
tracelessness constraint Cij
≡ δ ij hTT
ij are exactly preTT
served under the evolution: ∂t Cψ = 0, ∂t Cij
= 0 identically on solution.
Proof. Differentiate Cψ = πψ − ψ̇ in time: ∂t Cψ = π̇ψ − ψ̈.
The EL equation gives π̇ψ = ∇i (µϕi ) − ∂V /∂ψ = ψ̈ on solution, so ∂t Cψ ≡ 0. For TT-tracelessness: the projection
TT
operator Pij,kl
commutes with the wave operator □flat ,
so if δ ij hij = 0 at t = 0 it remains zero. Algebraically:
applying δ ij to the TT evolution equation gives a homogeneous equation for δ ij hij with zero source, which
by uniqueness of solutions of linear hyperbolic systems
is identically zero if it vanishes initially. The numerical
implementation verified ∥tr(h)∥ stays at machine zero
across all evolved steps.
c.

Convergence Kernel Theorem

Theorem AH.7 (Limber-Projected DFD Lensing Kernel). On the baryon-only apparent-amplitude construction,
which is not DFD’s growth carrier (proof, Step 6), the
weak-lensing convergence of a source distribution δb is
Z
κDFD (ℓ) =

dχ



1
3(H0early )2 χ g(χ)
ℓ+ 2
Ω
R
k
=
,
z;
δ
δb ,
b
DFD
b
χ
2c2
a(χ)

with response kernel RDFD (k, z; δb ) = Q(k, z; δb ) and
deep-MOND amplification QDFD ≈ 502.
Proof. Step 1.
The DFD optical metric ds̃2 =
2 2
2
2
−c dt /n + dx with n = eψ produces the lensing potential Φlens = −(c2 /2)δψ at linear order in δψ, identical in

functional form to the GR Newtonian potential. Step 2.
The
Born integral for convergence is standard: κ(n̂) =
R
dχ W (χ)∇2⊥ Φlens /c2 , where W (χ) = (χs − χ)/χs is the
lensing efficiency. Step 3. On FRW with App. AE Theorem AE.1 (āext = 0), the linearized DFD field equation
becomes k 2 δψk = (8πG/c2 )ρ̄b RDFD δb , with RDFD =
Q(k, z; δb ) and Q ≈ 502 in the deep-MOND regime where
|∇ψ| ≪ a∗ . Step 4. Substitute δψk into the lensingpotential integrand: ∇2⊥ Φlens /c2 = −(ℓ+ 12 )2 /(χ2 c2 )·Φlens ,
giving the angular-Fourier convolution. Step 5. Limber projection collapses the line-of-sight integral to the
displayed form. Step 6. Against the spherical-top-hat
R8 = 8 h−1 Mpc window the closure takes σ8 = 0.811
as its input and solves for QDFD (App. AE). This is
the optical amplification of a baryon-only contrast, the
apparent-amplitude construction of App. AE (no darkmatter halo entering the lensing kernel). DFD’s physical
carrier is the derived χ-matter (Ωχ h2 ≃ 0.12, ΛCDM-like;
App. AV), which sources the rest-mass channel on linear
scales, and the Rest-Mass Channel confines the baryons’
polarization to bound systems, so this amplification does
not act on the R8 scale and the lensing there follows the
matter power that χ and the baryons carry (App. GR of
the main volume).
Three pre-registered falsifiers of this baryon-only construction: direct Lyα σb (R8 , z), kSZ amplitude about 0.9
of LCDM’s, cross-survey F8 = 77.9 universality at ±10%.
They test the construction, not DFD’s growth carrier, the
cold χ relic with Qχ = 1 (main volume, App. AE and
App. GR).

d.

Cluster First-Principles Pipeline

The literature-bounded five-factor cluster budget does
not close the cluster shortfall (App. I); clusters need the
derived χ. The pipeline below is a fully specified zeroDOF first-principles ρb → (κ, σ-M, MHSE ) pipeline for the
µ-law side.
Theorem AH.8 (Universal µ in clusters, with χ). The
same interpolation function µ(x) = x/(1 √
+ x) that fits
SPARC galaxy rotation curves at a0 = 2 α cH0loc also
reproduces, without modification, (i) the cluster baryon
(gas) dynamics once the derived χ supplies the mass the µlaw leaves (about 0.84 of the cosmic mix inside r500 when
χ adds its Newtonian pull, and 1.21 ± 0.05 predicted from
the derived abundance under ΛCDM-like assembly, an
estimate; main volume, App. I), with the Bullet-Cluster
collisionless lensing offset carried by χ, and (ii) the multiscale-averaging effective coefficient µeff = µ·(1+0.39 fsub ).
Proof. (i) Bullet Cluster: χ adds its Newtonian pull
through the rest-mass channel, and the µ-law acts on
the baryons. During the merger the collisional gas is
shock-stripped and lags, while the collisionless χ (and
the galaxies) remain at the galaxy centroid and set the
lensing (κ) peak, displaced from the gas—reproducing the

6
observed offset by the same collisionless-mass mechanism
as ΛCDM. The µ-law on the baryons alone is insufficient:
at the two brightest cluster galaxies the baryonic field
is gN,b /a0 = 0.21–0.34 (boost ν = 2.3–2.7), and without χ DFD gives only 24–39% of the lensing mass inside
250 kpc; χ at 0.6–0.9 of the cosmic mix fits the lensing
map when it adds its Newtonian pull (1.1–1.2 under the
floor composition; main volume, App. I). The collisionless
carrier is supplied natively by the derived χ field; the
universal µ-function governs the baryon/gas dynamics.
(ii) Multi-scale averaging: substructure with subhalo
fraction fsub contributes an additive convex combination
µeff = (1 − fsub )µ + fsub µeff,sub . The constant 0.39 is the
angular average over substructure orientations. Numerics: an internal Python solver reproduces (i) qualitatively
at 643 resolution; full 16-cluster public-data run requires
AMR and is engineering-pending.

e.

Internal-Manifold Axiom Reduction

The reduction below uses the seven-axiom list of
App. AB ((V1), (V2), (V4), (V4′ ), (V5), (V7), (V10)). It
does not take the tangent-bundle gauge partition (3, 2, 1)
as an axiom (V*4), nor derive dim X = 7 from it: that
partition is topologically impossible on CP 2 × N 3 (the
tangent-obstruction lemma (App. AB, main volume)).
And π3 = Z does not follow from dim X = 7.
Theorem AH.9 (Reduced-Axiom Uniqueness of
CP 2 × S 3 ). The internal manifold X = CP 2 × S 3 is
uniquely determined (up to diffeomorphism) by the satisfiable axiom set of App. AB: compactness (V1), minimality
of dimension (V2), nondegenerate Kähler-times-frame
product structure (V4), parallelizable frame factor (V4′ ),
χtop (Mc ) = 3 (V5), π1 (Mg ) = 1 (V7), and positive internal Ricci curvature (V10).
Proof. Step 1. Corollary AB (dim-7; main volume):
(V10)+Kobayashi force Mc Fano and simply connected
with dimR Mc ≥ 4 (the only Fano candidates below
complex dimension 2 have χtop ∈ {1, 2} ̸= 3), and
(V4′ )+(V7) force dim Mg ≥ 3 (S 1 has π1 = Z; S 2 has
χtop = 2 ̸= 0; higher-genus surfaces have π1 ̸= 1). Minimality (V2), with the witness CP 2 × S 3 attaining 7,
forces dim X = 7 and the split 4 + 3. Step 2. Mc is
a del Pezzo surface with χtop = 3; Noether’s formula
leaves CP 2 as the unique entry (Lemma AB-Mc ; main
volume). Step 3. Mg is a closed orientable simply connected 3-manifold, hence S 3 by Perelman’s proof of the
Poincaré conjecture (Lemma AB-Mg ; main volume) —
no Lie-group hypothesis is needed. Step 4. On the
winner, kmax = χ(CP 2 , O(9) ⊕ O⊕5 ) = 55 + 5 = 60 by
Hirzebruch–Riemann–Roch, and Spinc -ness, π3 (Mg ) = Z,
and even mapping-torus dimension are derived properties
(Corollary AB-properties; main volume). Conclusion.
X = CP 2 × S 3 is uniquely determined; four properties
(dim X = 7, Spinc , π3 = Z, even dim TCP ) are theorems

of the axiom list, and kmax = 60 is a computed property
of the winner.
3.
a.

Statements

Padé Identity All-Orders Theorem and Firewall

The [1, 1] Padé identity of the main volume (App. AA)
extends to all orders, with an explicit firewall against
Yilmaz-type alternatives.
a. Notation. Throughout
this
subsubsection,
Pm,m (u) denotes the diagonal [m, m] Padé approximant
of eu , i.e. the unique rational function Nm (u)/Nm (−u)
with degree-m numerator and denominator whose Taylor
series matches eu through order u2m . By construction
Pm,m (u) ≈ eu , hence [Pm,m (u)]2 ≈ e2u .
Theorem AH.10 (Padé Hierarchy with Closed-Form
Coefficients). [Pm,m (u)]2 = exp(2u) + O(u2m+1 ) with explicit Taylor coefficients of the numerator polynomial
2m  
X
m (2m − k)! k
u .
Nm (u) =
k
(2m)!
k=0

The first divergence at O(u3 ) has coefficient ∆ = 3/2 −
4/3 = 1/6.
Proof. The diagonal Padé approximant has the standard
form Pm,m (u) = Nm (u)/Nm (−u) with closed-form nu
 (2m−k)! k
Pm
merator Nm (u) = k=0 m
k
(2m)! u (Bender–Orszag).
2
Squaring: [Pm,m (u)] = P
Nm (u)2 /Nm (−u)2 . Taylorexpand against exp(2u) = n (2u)n /n!. By construction
Pm,m (u) matches eu through order u2m , so the squared
difference vanishes to order u2m and the leading nonvanishing term is at u2m+1 . For m = 1: N1 (u) = 1 + u/2,
hence P1,1 (u) = (1 + u/2)/(1 − u/2) and [P1,1 (u)]2 =
(1 + u/2)2 /(1 − u/2)2 . Expand: (1 + u/2)2 = 1 + u + u2 /4;
(1 − u/2)−2 = 1 + u + 3u2 /4 + u3 /2 + · · · . Product: 1 + 2u + 2u2 + 3u3 /2 + · · · . Compare exp(2u) =
1 + 2u + 2u2 + 4u3 /3 + 2u4 /3 + · · · . Agreement through
u2 (= u2m for m = 1, as the general argument above
requires); first deviation at u3 (= u2m+1 ) with coefficient
∆ = 3/2−4/3 = 1/6, in agreement with the theorem statement and with the direct series comparison of App. AA
(which lists the u3 coefficients 3/2 vs. 4/3 explicitly). The
derivation in App. AA tracks the deviation through the
difference exp(2u) − [Pm,m (u)]2 to all orders, with leading
constant ∆ = 1/6, a property of the two functions.
Theorem AH.11 (m → ∞ Entire-Function Limit).
limm→∞ [Pm,m (u)]2 = exp(2u) uniformly on compact sets
in C.
Proof. Saff and Varga (1977) prove that diagonal Padé
approximants of exp(u) converge uniformly on compact
sets in C. Squaring is continuous, so the squared sequence
inherits uniform convergence. By Cauchy–Goursat, the
limit is the unique entire function matching exp(2u) at
all derivatives at u = 0.

7
Theorem AH.12 (Firewall vs. Yilmaz). The flat-R3
DFD geometry has no throat. The DFD field uψ (r) =
2GM/(c2 r) ≥ 0from the Lyapunov function W (X) =
√
√
X − 2 X + 2 ln 1 + X (main volume, App. GR), so
no exotic-matter signature appears. Yilmaz solutions arise
within GR as wormhole metrics requiring exotic stress–
energy.
Proof. At leading order DFD’s exterior solution is the
unique solution of the elliptic R
equation ∇2 ψ = 0 on R3 \
{0} with ψ → 0 at infinity and ∇ψ · dS = −8πGM/c2 .
This gives ψ = 2GM/(c2 r), defined and finite on all of
R3 \ {0}, with no double cover; with the parent strain’s σ
kept the full exterior is the Schwarzschild metric in DFD’s
coordinates, sourced by ordinary matter (App.
The
 WS).
√
√ 
Lyapunov function W (X) = X −2 X +2 ln 1 + X ≥
√
√
0 for all X ≥ 0 (W (0) = 0, W ′ (X) = X/(1 + X) ≥ 0)
confirms positivity of the trapping function. The Yilmaz
exponential metric, by contrast, when embedded in 4D
GR, requires exotic stress–energy (negative Ttt in some
patch) and admits a wormhole throat connecting two
asymptotically flat regions. The flat-R3 background of
DFD has no second region to connect to; topologically
the geometry is R4 \ {r = 0}, simply connected, no throat.
Detailed energy-condition analysis (App. AA) confirms
no NEC violation in DFD.

b.

Microsector Finiteness Theorems

Theorem AH.13 (Microsector Hilbert-Space Finiteness).
dim Hmicro = χ(CP 2 , E) = 60, computed by Hirzebruch–
Riemann–Roch on E = O(9) ⊕ O⊕5 .
Proof. The Hirzebruch–Riemann–Roch
formula gives
R
χ(CP 2 , F) = CP 2 ch(F) · Td(CP 2 ). For the rank-1 line
bundle O(d): ch = edh where h is the hyperplane class,
Td(CP 2 ) = (1 + 3h/2 + h2 ), giving χ(CP 2 , O(d)) =
(d + 1)(d + 2)/2. For d = 9: χ = 55. The trivial bundle
gives χ(O⊕5 ) = 5. Sum: χ = 55 + 5 = 60. Toeplitz quantization gives a finite-rank Hilbert space, not a regulated
infinite one.
Theorem AH.14 (kmax = 60 Index Computation at
the Derived Flux). kmax = 60 is the Atiyah–Singer index
of the canonical Spinc structure evaluated at the derived
q1 = 3 (main volume, App. F, the remark on the flux
quantum).
Proof. The isotropy charge 6Y fixes q1 = 3: a multiplet
of hypercharge Y lives in a homogeneous bundle of slope
3Y , and q1 = 6 would give the weak doublets an even
charge (main volume, App. F, the remark on the flux
quantum). The corresponding Spinc Dirac index on CP 2
is ind (DE ) = χ(CP 2 , E) by Atiyah–Singer; with E =
O(9)⊕O⊕5 (the a4 Seeley–DeWitt minimization argument
for this choice is asserted, not proven; the chain is that
of App. F’s independence remark, with q1 = 3 derived

and Ngen = 3, n = 5 inputs), this index is 60 as in
Theorem AH.13. The same integer is also reached as |A5 |,
as |E8 roots|/4 and as the shifted level of SU (2)58 ; these
routes are stated to give 60, not shown to reduce to one
another.
Theorem AH.15 (Asymptotic Freedom from Frame
Stiffness). The non-Abelian sector β-functions satisfy
β3 , β2 < 0, as in the Standard Model; the reading proposed here, not computed, attributes this to positive Ricci
of the projectivised mode-bundle base P (Vr ) = CP nr −1
via frame stiffness κr = nr κ0 .
Proof. The microsector Berry connection on P (Vr ) =
CP nr −1 — the projectivisation of the rank-nr block Vr ⊂
Hint of the internal mode bundle; this projective space
is the colour/weak fibre and is not the internal factor
Mc — has frame stiffness κr = nr κ0 , where κ0 is the
Fubini–Study scalar curvature. Positive Ricci gives a
positive-definite quadratic form on gauge fluctuations,
contributing −cr log(Λ/µ) to the running gauge coupling.
For nr = 3 (color SU(3)) and nr = 2 (weak SU(2)), the
contributions are read as exceeding the matter loops (not
computed), giving β3 , β2 < 0. For nr = 1 (hypercharge
U(1)), matter loops dominate, giving β1 > 0 but with
a topological cap at EUV where the running freezes (no
Landau pole, see Theorem AH.16).
Theorem AH.16 (No Landau Pole, No Trans-Planckian
Regime). α(E) → αUV at EUV . The running stops at the
topological cutoff scale because the Hilbert-space dimension
caps the flow.
Proof. The renormalization-group flow involves integrating
√ out modes with |p| < E. Once E exceeds Λtop =
αMP , all 60 microsector modes are in the spectrum
and no further modes can be integrated out (since
dim Hmicro = 60 is finite). The running freezes at
α(EUV ) = αUV . By Theorem AH.13, this is finite. By
the Abelian beta function with β1 > 0 but capped flow,
no Landau pole exists.
c.

Higgs Hierarchy

Theorem AH.17 (Higgs √
Vacuum Expectation Value:
the Exponent). v = MP α8 2π = 246.09 GeV (0.05% to
PDG), with the exponent 8 derived by dimension
√ count
(8 = dim(CP 2 × S 3 ) + 1 = 7 + 1), the prefactor 2π fitted
(main volume, App. Z), and the loop-counting (α2 )4 and
twist-bundle Chern routes recorded as consistency restatements (not independent confirmations, per App. AY).
Proof. Primary route (dimension counting).
dim X + 1 = 7 + 1 = 8, where dim X = 7 is Theorem AH.9 (dim CP 2 = 4 plus dim S 3 = 3) and the +1 is
the singlet/radial direction normal to the gauge bundle.
This is the single derivation that fixes the exponent.
Consistency restatement A (loop counting).
Each gauge-coupling 2-loop factor contributes α2 ; the

8
four microsector factors (gauge-Higgs, Higgs-Yukawa,
gauge-Yukawa cross, self-Higgs) give (α2 )4 = α8 . Consistency restatement B (twist-bundle Chern).
The bundle E = O(9) ⊕ O⊕5 has
R total Chern class
c(E) = (1 + 9h)(1)5 = 1 + 9h, with CP 2 c1 (E)2 = 81, giving Chern character contribution ∝ 9 + (−1) = 8 via the
Atiyah–Singer correction. Not independent. All three
agree on 8, but the loop-counting and Chern routes are
consistency restatements, not independent confirmations:
App. AY records the cross-sector lattice identities of this
type as tautological (equivalent restatements of a single
α-exponent, carrying no information beyond it). The
dimension count alone is the derivation.
√ Numerics. v =
1.220890 × 1019 GeV × (1/137.036)8 × 2π = 246.09 GeV,
vs. PDG 246.22 GeV, error 0.05%.
Remark AH.18 (The Planck-scale boundary). Imposed
at MP , λH = 1/8 runs to λH (v) = 0.174 under two-loop
Standard-Model running and gives mH = 150 ± 2 GeV,
far above the measured value, so the Planck scale is
not the scale at which the conjecture holds. The main
volume states λH = 1/8 as a conjecture at the electroweak
matching scale (App. Z): imposed at the top-quark mass it
gives mH = 124.63±0.15 GeV, 0.40% below the measured
125.13 ± 0.11 GeV (PDG 2026; computed), and λ = 1/8
holds exactly at 182 ± 4 GeV.

d.

cT = c Structural Theorem

Theorem AH.19 (Parent-Tensor Decomposition). ψ
(trace) and hTT
ij (TT) are the two O(3)-irreducible components of the same zero-mode metric perturbation on
K = CP 2 × S 3 .
Proof. The zero-mode tensor parent on K decomposes
under Iso(K) ⊃ SO(3) via Lichnerowicz analysis. Koiso
(1980) classified scalar zero-modes on Kähler manifolds:
only χ(CP 2 ) = 3 contributes, of which the trace and
squashing are the two components. Higuchi (1987) classified S 3 zero-modes: the trace lifts to S 3 trivially. Combined: K admits a single 4D scalar ψ (trace) and a single
TT-spin-2 mode hij . On the truncated microsector there
is no size or shape modulus (main volume, App. O). Oneform components have b1 (K) = 0 on CP 2 × S 3 and are
excluded. A further harmonic channel is the three-form:
by Künneth b3 (K) = b0 (CP 2 ) b3 (S 3 ) = 1, the bi-invariant
Cartan three-form on S 3 = SU (2), which is not a metric
mode and so does not affect cT = c (Theorem AH.20); its
4D reduction on the three-cycle is the pseudoscalar matter
field χ (the derived cold dark matter; App. AV).
Theorem AH.20 (cT = c from O(3) Block-Diagonality). O(3) block-diagonality of the principal symbol forbids
(∂ψ)(∂hTT ) derivative mixing. Light and gravitational
waves share one metric (App. WS).
Proof. The propagator matrix M(p) of (ψ, hTT
ij ) in DFD
is O(3)-covariant: ψ is a scalar, hTT is spin-2 with helicity

±2. By Schur’s lemma, no O(3)-covariant matrix can mix
scalar and spin-2 components; the propagator is blockdiagonal. On the background the tensor block satisfies
(∂t2 − c2 ∇2 ) hTT
ij = 0, giving phase velocity c; the scalar
is an instantaneous constraint that does not radiate (Axiom WS.1). GW170817 is satisfied by construction.
Block-diagonality forbids the bilinear mixing only; the
tensor cone itself is fixed by local relativity at constant ψ0
and equals the optical cone c e−ψ (main volume, Sec. V A,
Remark V.1; App. WS).

e.

√
a0 = 2 α cH0 Derivation Chain

3
Theorem
√ AH.21 (MOND Scale from S Composition).
a0 = 2 α cH0 from the unique microsector composition
law on S 3 (Theorem N.14 of App. N).

Proof. Four-step chain following App. N. Step 1 (Witten partition exponent). The S 3 partition function
with Toeplitz cap at kmax = 60 gives a discrete spec1/2
trum {λk }60
. Step 2
k=1 with leading scale λ1 ∝ α
(composition law). Lemma N.6 [App. N]: under S 3
saturation-union, the response composes as ψ12 = ψ1 + ψ2
for nonoverlapping sources, fixing the interpolation function µ to the exponential form 1 − e−cψ . Step 3
(weak-response form). Theorem N.8: with the weakresponse level response keff = k0 (1 + s) this becomes
µ(s) = s/(1 + s). Step 4 (variational stationarity).
The variational
principle δΞ/δψ = 0 at Ξ∗ = 3/2 gives
√
a0 = 2 α cH0loc = 1.1966 × 10−10 m/s2 on bound-matter
clocks (main volume, App. ZS) and 1.1208 × 10−10 on the
early clock. Against a0 = (1.20±0.02stat ±0.24sys )×10−10 ,
the bound-matter value lies within the systematic error;
the early-clock value is −3.96σ statistical and −0.33σ
systematic.
H0 caveat. The C = 1, H0 = 72.09 clock is excluded
as the early clock and is the bound-matter clock of the
two-clock postulate (main volume, App. ZS), so bound
systems read a0 = 1.1966 × 10−10 m/s2 , consistent with
SPARC within
its systematic error. Its closed form is
p
a0 = 2α29 c7 /Gℏ, the C = 1 value of the clock identity,
with no measured coefficient; on the early clock, where
5
57
GℏH02 /c
Cα
the same law gives
p with C measured, −10
√ = 29
a0 = 2 C α
c7 /Gℏ √
= 1.1208 × 10
m/s2 . Independently, the prefactor 2 α = 0.171 differs by 7.3% from
the empirical Milgrom coincidence a0 ≃ cH0 /2π (0.159):
the order a0 ∼ cH0 is robust and mass-independent
(turnaround/virialization),
whereas the exact coefficient
√
2 α is the DFD-specific claim, resting on the Ξ∗ = √
3/2
stationarity step (App. N). That this step forces 2 α,
rather than merely admitting it within the generic O(1)
spherical-collapse prefactor band (which also contains
1/2π and 1/6), is the key item not yet shown.

9
f.

Topological H0 and Cosmological Constant

Theorem AH.22 (H0 from Topological Invariant).
GℏH02 /c5 = C α57 . The exponent 57 = kmax − Ngen
is forced by primed-determinant scaling (Thm. AH.23);
the dimensionless coefficient C is not derivable from the
printed axioms and is measured, C = 0.87784 ± 0.00266
(BAO+θ∗ ; plik-lite 0.8772), and the cross-channel earlyclock value is H0early = 67.52 ± 0.09 km/s/Mpc; boundmatter clocks read C = 1 under the two-clock postulate
(main volume, App. ZS).
=
Proof. By Theorem AH.23 below, det′ (g∆)
g kmax −Ngen det′ ∆ = g 57 det′ ∆ on the dynamical
microsector with g = α. The dimensionless invariant constructed from G, ℏ, H0 , c with weight zero in
mass and length is uniquely GℏH02 /c5 (other dimensionless products with these constants miss by ≥ 17
decades). Equating to the determinant ratio fixes
the exponent: GℏH02 /c5 = C α57 , with the dimensionless C √left undetermined by this
p argument. Solving,
H0 = C α28.5 /tP , where tP = Gℏ/c5 is the Planck
time. At the measured C = 0.87784 ± 0.00266 (the
clock-credit remark, App. AP of the main volume) this
gives H0early = 67.54 ± 0.10 km/s/Mpc; the adopted
cross-channel value is 67.52 ± 0.09. A C = 1 reading
returns H0 = 72.0902; that reading is excluded at ∼46σ
as the early clock by the CMB-anchored channels; on
bound-matter clocks it is the local value H0loc = 72.09 of
the two-clock postulate (main volume, App. ZS).
Theorem AH.23 (Cosmological Constant via Determinant Scaling). 57 = kmax − Ngen = 60 − 3 is forced by
primed-determinant scaling (Lemma O.1) via per-mode
α-cancellation (Lemma O.5).
Proof. Lemma O.1 [App. O]: det′ (g∆) = g nnz det′ ∆
where nnz = dim ker⊥ ∆ = dim Hmicro − dim ker ∆.
Lemma O.5 [App. O]: per-mode α-cancellation eliminates
eigenvalue dependence; only the count nnz enters. By Theorem AH.13 dim Hmicro = 60, and by chiral generation
counting dim ker ∆ = Ngen = 3. Hence nnz = 57. The
current dimensional invariant is Cα57 = 1.393 × 10−122 ;
the determinant proves the exponent, while the residual
coefficient C = 0.87784 ± 0.00266 is measured. The equality 1.586 × 10−122 = α57 is the C = 1 value, excluded as
the early clock and held by bound-matter clocks under
the two-clock postulate (main volume, App. ZS).

g.

expectation ⟨Ψ|ρ̂|Ψ⟩, and the µ-monotone elliptic PDE
admits a unique weak ψ for any L2loc source. There is no
manifold branching.
Proof. DFD’s field equation ∇ · [µ(|∇ψ|/a⋆ )∇ψ] =
−(8πG/c2 )ρ is a strictly µ-monotone elliptic PDE
(App. U). For a quantum source state |Ψ⟩, the source
term
is the c-number expectation ρ → ⟨Ψ|ρ̂|Ψ⟩ =
P
p
ρ
i
i (probability-weighted sum over branches). By
i
the Browder–Minty theorem applied to the strictly monotone operator, the elliptic equation has a unique weak
solution ψ ∈ H 1 for any L2loc source. Hence there is a
single ψ-field, not a branched superposition of ψ-fields
per Penrose-paradox premise (P1). Penrose’s premise
that “each branch sources its own metric” fails: in DFD
the metric is sourced by the weighted probability density,
just like any classical field sourced by a quantum charge
density. The paradox does not arise.

Theorem AH.25 (Quantum-Operator ψ̂). In the hyperbolic relaxation form of its field equation, ψ̂ admits
canonical commutator [ψ̂(x), π̂ψ (y)] = iℏ δ 3 (x − y) and
is a flat-background scalar QFT (dimensionless field with
the Planck-scale stiffness c4 /32πG in front of its kinetic
term, no curvature counterterms). Under Axiom E of the
main volume the local sector has no temporal kinetic term,
so there ψ̂ has no independent canonical momentum and
is fixed by its source.
Proof.
The
DFD
scalar
action
Sψ
=
R
dt d3 x (a20 /8πG) W (|∇ψ|2 /a2⋆ ) of the main volume
(no potential term in ψ; Theorem AH.136, sub-(i))
on flat R3,1 has dimensionless ψ, whose kinetic term
in physical units is (c4 /32πG)(∂ψ)2 = (M̄P2 /4)(∂ψ)2 :
it is not marginal (for a dimensionless field that
needs z = 3 anisotropic
√ scaling,√which DFD does not
have), and ϕ = M̄P ψ/ 2 = ψ/ 16πG is an ordinary
√
canonical scalar (the decay constant F = M̄P / 2
of the main volume, App. FK). Standard canonical
quantization of the relaxation form on a Cauchy slice
gives the displayed CCR. Power counting in ϕ: the
kinetic term (∂ϕ)2 has dimension 4; a potential term
V (ψ), absent in DFD, would also be dimension 4;
interactions are non-renormalizable in the standard
sense. The Hilbert space is F DFD = F ψ ⊗ Hmicro with
dim Hmicro = 60, finite. The two descriptions (flat-scalar
QFT and microsector-augmented) agree to O(ℓP /L) on
observables.

Penrose Superposition Paradox Dissolution

Theorem AH.24 (Penrose Paradox Dissolution). In
DFD the optical metric n = eψ is sourced by the c-number

h.

Fine-Structure Constant Closed-Form Identity

Theorem AH.26 (α−1 Closed-Form).


3/2
+3
7
α−1 = π24 · Tr(Y 2 ) · kmax · kkmax
· 1 + 80·4095
= 137.03599985,
max +4

7
1
2
80·4095 = 13 kmax

exactly

10
with residual −0.005 ppm vs. CODATA 2022. The closed form is exact under two stated choices (main volume, the
convention-locked closed form for α), and the bracket’s coefficient 1/13 is a stated rational whose provenance is open.
2
Proof. The prefactor π 3/2 /24 is the heat-kernel normalization of the exact
√ construction: the ζ(0) of the CP charge trace,
−1/24 per unit charge squared, times the terminating heat trace ( π/4)Λ3 of the ρ-shifted Casimir on S 3 = SU (2),
with the heat-kernel regulator. The factor k(k + 3)/(k + 4) at k = kmax = 60 (Theorem AH.14) combines the exact
CP 2 charge-trace ratio (k + 3)/k with the Toeplitz dimension k + 4. The hypercharge trace is Tr(Y 2 ) = 10, three
generations of 10/3; its identification with 1/e2 rather than with the hypercharge coupling is not derived (main
2
volume, the convention-locked closed form for α). The bracket [1 + 7/(80 · 4095)] = 1 + 1/(13kmax
) has a stated
−1
rational coefficient. Numerical evaluation gives α = 137.03599985, with residual −0.005 ppm vs. CODATA 2022
137.035999177(21); the integers alone give 137.03307.

i.

Charged Fermion Mass Formula Tightening

j.

Neutrino Sector: Normal Ordering and δCP = −π/2

Theorem√AH.27 (Mass Formula Decomposition). mf =
Af αnf v/ 2, with sector exponents nf ∈ {0, 1.0, 1.5, 2.5}
derived from Cayley-graph hop distance, and prefactors Af
traced to four irreducible blocks: Kd , Ku (CP 2 kernels);
Qd (QCD); Dℓ (Dirac); G (generation).

Theorem AH.28 (Mass Ordering and CP Phase). On
the adopted neutrino branch, normal ordering is forced.
δCP = −π/2 from the µ–τ reflection, a structural postulate (main volume, App. FK), with the sign set by the
Fubini–Study orientation.

Proof. The Yukawa operator factorizes as Yf =
X+1
Af hdim
via the Berry-bundle overlap integral on CP 2 .
f
The sector exponent nf counts the hop distance on the
Cayley graph from the Higgs vertex to the fermion vertex:
n = 0 for top and bottom, n = 1 for charm and tau,
n = 1.5 for strange and muon, n = 2.5 for up, down
and electron (main volume, App. FM). The prefactor Af
decomposes as

Proof. The DFD neutrino mass matrix derived from microsector (App. X) yields branch-B exponents k = α−3/11 ,
r = α−7/20 with absolute scale m3 = (14/13)πMP α14 .
Both k, r > 1, so the mass ratios m2 /m1 = k, m3 /m2 = r
enforce m1 < m2 < m3 (normal). Inverted ordering
m3 < m1 ≈ m2 is structurally impossible. The CP
phase δCP is the imaginary part of the PMNS Jarlskog
invariant, computed from the complex structure on CP 2
(Fubini–Study Kähler form ωFS ). The phase is fixed by
the µ–τ (S2 ) reflection symmetry S h S = h∗ acting on the
neutrino mass matrix: this antiunitary symmetry forces
θ23 = π/4 and | sin δCP | = 1 simultaneously at all orders,
and the Fubini–Study orientation selects the sign, giving
δCP = −π/2. Evaluated at the DFD-locked mixing angles
(sin2 θ13 = 3α, θ23 = π/4, trimaximal sin2 θ12 = 0.3184,
from |Ue1 |2 = 2/3) with sin δCP = −1, the leptonic Jarl2
skog
√ invariant
√ √is JPMNS = s12 c12 s23 c23 s13 c13 sin δCP =
−( 2/6) 3α 1 − 9α = −0.0337 (|JPMNS | = 0.03371, as
in the main volume, App. AY); the maximal-angle prefactor 1/8 is not the physical invariant. This is consistent
with current T2K/NOvA hints. Combined predictions:
∆m221 = 7.48 × 10−5 eV2 , ∆m231 = 2.51 × 10−3 eV2 ,
NuFIT 6.0 χ2 = 0.025, p = 0.99.

Af = Kf · Qf · Df · Gf ,
where each block is fixed by symmetry: Kd = J3 , Ku =
I4 from S3 /O(4) symmetry on the kernel space, with
democratic coupling for Kd (Lemma K.2/K.3); Qd =
diag(1, 6/7, 1/42) from
√ the QCD anomaly coefficient b0 =
7; Dℓ = diag(1, 1, 2) from Dirac normalization on the
lepton sector; G = diag(2/3, 1, 1) with G[1, 1] = 2/3 from
primed microsector trace (Theorem K.4). All 9 charged
fermion masses are computed at this stage with mean error
1.42% and maximum error below 4%; the full per-fermion
table is supplied via the unified Python reproducibility
script which evaluates all 34 Standard-Model continuous
parameters at 50-digit precision (the running factors of
Theorem AH.200 are not Standard-Model running, so no
sharper mean error is claimed). Scope caveat: the α-power
exponent ladder nf and the equal-exponent ratio structure
are derived ; the absolute charged-fermion spectrum below
the top, however, additionally imports seven data-selected
prefactors, one for each species other than the top and
the charm (main volume, App. FK and App. FM). The
reproducibility script therefore reproduces the 34 values
given those selected/fitted inputs; the quoted mean errors
(1.42% and 0.61%; the 0.082% of Theorem AH.200 does
not stand) measure the fit quality of that structure, not a
from-first-principles derivation of all 34 parameters.

4.

New Theoretical Territories
a.

Topological Pre-Inflation

Theorem AH.29 (Lichnerowicz Rigidity Excludes
Slow-Roll Inflaton). On CP 2 ×S 3 the Lichnerowicz Laplacian leaves no slow-roll scalar: on the truncated microsector there is no size or shape modulus, and on a continuous
internal space the only scalar moduli are the overall volume and the squashing, whose curvature potential has no
slow-roll direction.

11
Proof. The Lichnerowicz Laplacian on a compact Einstein
manifold acts on transverse-traceless symmetric 2-tensors.
On CP 2 (Koiso 1980) the only zero modes are conformal
Killing forms; squashing of the Kähler–Einstein metric
is the unique non-trivial zero mode. On S 3 (Higuchi
1987) the zero modes are the Killing forms; b1 (S 3 ) = 0.
Combining via K = CP 2 × S 3 with b1 (K) = 0, the only
scalar moduli of a continuous K are its overall volume
√
and the squashing modulus: the shape sits at τ∗ = 1/ 3,
a minimum of the curvature potential for the connection
Laplacian, and the volume has no critical point at this order, a runaway rather than a slow roll. The Kähler radius
is a modulus of K, not the field ψ, whose homogeneous
mode is set by the Friedmann law (main volume). On the
truncated microsector none of these moduli exists. No
slow-roll-compatible scalar exists.
Remark AH.30 (Status note on TPI). The Topological Pre-Inflation construction is refuted as a
DFD theorem. The {1, 3, 12, 30, 60} staircase appearing in the construction does not appear in the
main volume; the canonical cohomological catalog is
{3, 7, 8, 13, 19, 31, 49, 57, 60, 108, 137}. Lichnerowicz rigidity (above) excludes graviton-mode inflaton candidates
but does not exclude an external scalar inflaton field added
to DFD. Independent recomputation finds ns indeterminate (closest natural prescription = scale-invariant 1.000,
ruled out at > 8σ), r indeterminate. TPI is a conjectured extension beyond DFD’s stated scope, not a
derived prediction.
Theorem AH.31 (TPI Predictions, conditional on the
construction’s premises). Conditional on the staircase
construction, the predicted observables are: ns = 0.964 ±
0.003, As = (2.10 ± 0.08) × 10−9 , r = (4.0 ± 1.5) × 10−3 ,
TRH ≈ 5 × 1012 GeV.
Proof. The construction posits a discrete sequence of
effective dimensions {1, 3, 12, 30, 60} each contributing
∆Ne ≈ 12 e-folds via vacuum-energy-driven expansion.
The spectral tilt is determined by the rate of staircase
climbing: at each step, H ∝ α−1/2 scales, and the slowroll parameter ϵ ∼ 1/Ne2 gives ns = 1 − 2ϵ − 4η = 0.964
at Ne = 60. As from the COBE
normalization at the
√
staircase’s Hubble rate H ∼ αMP . r = 16ϵ ≈ 4 × 10−3 .
TRH from Γinf = MP α6 ⇒ TRH = MP α3 ≈ 5 × 1012 GeV.
Note: this is conditional on the staircase premise, which
is challenged by the status note above. Its r ≈ 4 × 10−3
also needs a freeze-out at H ≈ 1.6 × 1013 GeV, which
erases the χ relic (Theorem AH.40), so the construction
is incompatible with DFD’s dark sector.

b.

Baryogenesis from S 3 Topology

Theorem AH.32 (Shared-π3 Double Duty). π3 (S 3 ) = Z
of the internal winding is the same group as π3 (SU (2)) =
Z for sphaleron winding.

Proof. S 3 = SU (2) as a topological space and Lie group.
The third homotopy group π3 (S 3 ) = Z (Hopf invariant)
is therefore identical to π3 (SU (2)) = Z (winding number
on the SU(2) gauge bundle). The same group Z, on two
different three-spheres (main volume, App. F), classifies
both S 3 -twists in the DFD microsector (Theorem F.26)
and sphaleron transitions in electroweak gauge theory
(Manton 1983).
Thermal leptogenesis does not give ηB in DFD: for
the minimal DFD lepton sector the Dirac and Majorana
neutrino kernels are real and determinant-protected, so
the standard thermal-leptogenesis source
X 

εi ∝
Im (Yν† Yν )2ij f (Mj2 /Mi2 )
j̸=i

vanishes identically: Yν† Yν ∈ R ⇒ Im[(Yν† Yν )2ij ] = 0 ⇒
minimal DFD
εi = 0, hence ηB
= 0. This was verified two independent ways: the unflavored sum Im Tr[(Yν† Yν )2 ] = 0
(trace of a Hermitian square), and the resonant regulator
Mi2 − Mj2 = 0 for the exactly-degenerate heavy spectrum
of Theorem P.4. The same real-kernel structure that enters the strong-CP result (θ̄ = 0 at tree level, loop-bounded,
with θbare = 0 fixed by the spectral action given that nothing is added to it for the gauge fields) removes the CP-odd
invariant required for leptogenesis — DFD’s strong-CP
virtue and its baryogenesis are in structural tension. A
nonzero asymmetry would require two further DFD-SD
ingredients, neither presently derived: a geometrically
forced leptonic determinant-orthogonal Berry offset (its
magnitude, not merely its existence and tree-level, loopbounded strong-CP safety, which follow given the stated
premises) and a forced heavy-neutrino splitting ∆MR
(Theorem P.4 fixes only the determinant/geometric-mean
eigenvalue, not the individual masses). This thermalleptogenesis no-go is not the final word: the internal axial
Berry-holonomy channel of Theorem AH.33 below derives
the α4 power of the asymmetry (its bias current is B−L,
and with the stated carrier it gives about (2–3)×10−9 , three
to five times Planck’s value); the discussion in this note
establishes only that the thermal source vanishes. Given
the identified Majorana scale (main volume, App. P),
what DFD fixes here: MR = 4.744 × 1012 GeV and the
strong-washout efficiency κf = 4.87×10−3 (effective mass
m̃1 = m3 = 50.16 meV, K = 46.9; numerically-solved
Boltzmann) of the strongest of three washout channels,
K = (2.19, 8.38, 46.9) for the degenerate heavy spectrum
with real kernels (computed) — these set the scale and
washout, but not the CP asymmetry, which is the open
quantity.
Rolling-phase (spontaneous) route. A third mechanism, distinct from thermal leptogenesis and from heavysplitting, was examined: a time-dependent determinantorthogonal Majorana Berry phase MR → MR eiΘL (ψ) ,
which would generate a spontaneous lepton chemical potential µL = 12 Θ̇L on the active-lepton current via field
redefinition. This route is attractive because it requires
neither a nonzero εi nor a heavy splitting. The mechanism

12
itself is sound: because NR is Majorana (N̄R γ µ NR = 0),
rephasing carries the phase through the Dirac Yukawa
onto the nonzero current jLµ = L̄γ µ L, giving µL = 12 Θ̇L
(standard Majoron field redefinition), and the resulting
B − L is reprocessed by sphalerons consistently with proton stability (Thm F.26, Step 2). Define the dimensionless
Berry-clock invariant QL ≡ Θ̇L /(2H). Using the locked
inputs (MR = 4.744 × 1012 GeV, TL ≃ 2.0 × 1012 GeV,
κf = 4.87 × 10−3 , asph = 28/79, g∗ = 106.75), the conditional asymmetry is
ηB ≃ 1.08 × 10−9 (QL /α),
so the observed ηB would require QL = 0.566 α. This
does not close baryogenesis, for three structural reasons
established under the no-fitting rule (no fitting to ηB ).
(i) The forced phase is static, not rolling. The geometry forces only the static determinant-orthogonal Berry
offset of the quark prototype (Postulate Y.10, App. AO):
one conjugation-odd coordinate over the fixed internal
space CP2 × S 3 , with no ψ-argument. A static ΘL is a
pure rephasing (µL = 0), recovering the εi = 0 no-go
above. (ii) The ψ-promotion is obstructed. The
FRW flat-direction theorem (Thm AE.1, App. AE) renders the homogeneous cosmological ψ̇ non-dynamical—
the very drive the mechanism needs. (iii) The slope
is free. The roll-rate qL = 12 dΘL /dψ has no selection
rule (strictly weaker than the CKM apex Postulate E.1,
which at least admits the convex Euler-projection action);
any small rational {4/7, 13/23, 30/53, . . . } reproduces
0.566 α to ≲ 1%, the signature of a fit. A more sophisticated dressing—identifying QL = α/2 from a Spinc
Pfaffian-line “holonomy per generation”—fails for an independent reason: the proposed 12 = (Ngen /2)/Ngen is
an Ngen -independent algebraic identity (it returns 12 for
any generation count, so the index computation plays no
role), the flux-to-rate normalization λψ = α is unforced
(no such rule exists in the geometry, and the family “flux”
it halves is a degree-0 fiber index over the fixed internal
manifold, not a degree-2 Berry class over ψ-space), and
the resulting ηB = 5.4 × 10−10 is excluded by the CMB
(6.13 × 10−10 ± 0.7%) at ∼ 17σ. Moreover, taking the
clock-tied ψ̇ ∼ H(TL ) gives ηB ∼ 5 × 10−12 for qL ∼ α—
two orders too small ; the closure value tacitly assumes
ψ̇/T about 100× larger than H/T . DFD reduces baryogenesis to one determinant-line Berry-clock invariant QL ;
the geometry forces it static (no asymmetry), the rolling
spacetime-ψ promotion is cosmologically obstructed and
the slope is unforced, and the estimate undershoots by
two orders. The spacetime-ψ rolling route therefore does
not close baryogenesis; the internal axial Berry-holonomy
route of Theorem AH.33 below evades all three obstructions and derives the α4 power of |ηB | and its B−L bias;
with the stated carrier it gives ηB of about (2–3) × 10−9
when every step is pinned to the print, three to five times
Planck’s value, and its sign rests on an adopted axiom
(Theorem AH.33).
We do not adopt a baryogenesis branch axiom. Generating ε1 ̸= 0 requires injecting a CP-odd kernel defor-

mation that DFD’s real-kernel/determinant-protection
structure forbids — the same structure that keeps
arg det(Mu Md ) = 0 — and the resulting ηB would be
a fit, not a prediction. That is non-DFD content (a
CP source the geometry rules out), not an extension of
minimal DFD. The ε1 = 0 no-go stands.
Internal axial Berry-holonomy channel: the α4
power. The three obstructions to the rolling-phase route
above all attach the clock to the spacetime field ψ. Relocating the clock to the internal determinant line of the
chiral kernel—a connection on the CP2 × S 3 fermion bundle, not a ∇ψ scalar—removes (ii) and (iii) and reduces
(i) to a single named insertion. This is not the forbidden
ε1 ̸= 0 branch axiom (no CP-odd kernel is injected; the
real-kernel/θ̄ = 0 protection is untouched): the asymmetry is a CPT-violating chemical-potential bias. We record
the resulting theorem.
Theorem AH.33 (Internal axial Berry-holonomy baryogenesis: α4 power and B−L bias derived). Let the determinant line of the chiral fermion kernel carry, on the
retarded/expanding branch, the flat internal connection
α
A5 = Q d ln A,
Q= √ ,
π
so that in absolute time its temporal holonomy is a (B−L)
chemical potential µB−L = A50 = QH. Then, with the
3
locked seesaw scale MR = M
√P α and the conventional
freeze-out point TL = MR / 2π (a convention, not the
freeze-out of a computed process; proof, step (4)), the
sphaleron-reprocessed entropy-normalized asymmetry is
nB
YB ≡
= asph YB−L ,
s
15 gL  H 
YB−L =
Q,
YB = 0.0292 α4 ,
4π 2 g∗ T TL
and the photon-normalized baryon-to-photon ratio is obtained by the entropy-to-photon conversion s/nγ = 7.04,
s
15 gL  H 
base
ηB
=
YB = 7.04 asph 2
Q
nγ
4π g∗ T TL
= 0.2057 α4 ≃ 5.83 × 10−10
[base (deposit-free)],
which the per-e-fold transport deposit completes to
ηB = 0.2057 α4 (1 + 2πα) = 6.1007 × 10−10
(Thm. AH.36). These are the values at the conventional
√
freeze-out point with the Gaussian carrier Q = α/ π,
which is a premise the print cannot supply (proof, step (1)).
The α4 power of |ηB | is derived, and so is the current
that carries the bias, B−L (proof, step (3)). The printed
coupling fixes the object the transport turns, the determinant line, and its B−L charge, but not its turn per
flux step, which is stated: the stated carrier gives each
generation a full B−L turn per flux step; with it, on the
computed freeze-out, the mechanism gives ηB of about (2–
3) × 10−9 pinned to the print, three to five times Planck’s
value (proof, step (4)), so the printed transport overshoots.
The sign (matter vs. antimatter) is fixed by the expansion

13
orientation of Axiom AH.35, a CP-odd datum carried
by A5 on the determinant line — adopted, not derived
(the baryon winding itself carries zero information about
sign(w0 )).
Proof. (1) The width: its α1 power is derived, its normalization is not. The clock is the determinant line of
the chiral Dirac kernel—a complex determinant line (a
complex line bundle), not a real scalar oscillator. DFD’s
determinant-scaling lemma (the same that fixes the exponent of H0 , MR = MP α3 , and the Higgs/fermion α-tower)
assigns one complex Gaussian mode the partition-function
ratio
R −(λ/α)|z|2 2
e
d z
Zα
= α,
= CR −λ|z|2 2
Z1
d
z
e
C
i.e. scale parameter
√ α (a standalone real scalar would give
the half-power α; the complex line gives the full α). The
two orientation branches average to zero, ⟨q⟩± = 0; the
retarded branch restricts to the oriented half-line q ≥ 0,
giving the folded one-branch mean
Z ∞
2
2
2
Q = ⟨q⟩+ =
q √ e−q /α dq
α
π
0
p
√ 
α 
=√
= σ 2/π with σ = α/ 2 .
π
No datum is fitted, but the normalization is not fixed.
Reading α as the absolute scale of this Gaussian is a
premise, not a result: the Gaussian ratio fixes the α1
power and not the coordinate in which the folded mean
is taken, because it is blind to the normalization of the
coordinate it integrates (the flux-period lemma, part (ii),
of App. AV). Nor can a Gaussian carrier supply the
bias: a Gaussian vacuum carries no Berry phase, and
a single complex Gaussian mode
√ of ratio α has a normally
distributed coordinate of size α, not α (computed). The
realizable carrier is the determinant line’s holonomy over
one transport period, a whole number of turns, Q =
2πα nw /|cX |, with nw the line’s windings per e-fold and
cX its B−L charge (step (4)).
(2) The internal holonomy escapes the ψ-clock subtraction. A5 is a connection on the internal CP2 × S 3 fermion
bundle, not a ∇ψ scalar. The absolute-time segmentidentification (App. Q) and the FRW flat-direction theorem (Thm AE.1, App. AE) quantify only over ∇ψ/screenflow data and the ψ-action; they leave an internal axial
connection untouched. A chemical potential is a flat
(FA5 = 0) temporal connection whose
physical
content is
H
H
the thermal/cosmic Wilson line A5 = Q d ln A, not its
curvature; it is therefore not removed by the static-offset
reasoning (obstruction (i)) nor by the ψ̇ non-dynamics
(obstruction (ii)). Its drive is the Friedmann H at TL ,
fixed by the radiation density, which the flat-direction
theorems do not zero.
(3) The CP-odd datum. A chemical potential for B−L
is CP-odd as well as CPT-odd, and the S 3 flux step is
CP-even under the CP of App. L of the main volume,
so the only CP-odd datum that can drive a net B−L is

the adopted orientation of the transport (Axiom AH.35),
carried by the connection A5 on the determinant line,
which DFD’s antilinear CP lift conjugates, while the flux
step only sets one transport period per e-fold; the sign is
adopted, not derived. The current is B−L, not the axial
current J5 : a current with an axial quark component is
QCD-anomalous and, coupled to the flat, never-ending
A5 , would shift θ̄ by 12Q per e-fold, at least 109 times
the neutron-EDM bound even over the 2.3 e-folds in
which lepton number is violated; lepton-only currents
are P-odd or bias processes that conserve B−L; and the
charges conserved by all 22 B−L-conserving interactions
are exactly B−L and hypercharge (computed). The realkernel / even-dimensional η-invariant protection (App. L)
forbids CP-odd operators of the θ-term type at tree level
(θ̄ = 0, with θbare = 0 fixed by the spectral action given
that nothing is added to it for the gauge fields; εi = 0)
but not a CPT-violating background, so strong CP is
preserved and the εi = 0 no-go is irrelevant to this channel.
(4) Magnitude. With gL = 6, g∗ = 106.75, √
asph =
√
28/79, (H/T ) = 1.66 g∗ T /MP , and TL = MR / 2π =
√
MP α3 / 2π,
15 · 6
28
base
ηB
· 2
= 7.04 ·
79 4π · 106.75
√
α
α3
· 1.66 106.75 · √ · √
π
2π
= 0.2057 α4 = 5.83 × 10−10
[base (deposit-free)],
base
ηB = ηB
(1 + 2πα) = 6.1007 × 10−10

(deposit factor: Thm. AH.36).
4

The power α = α3 [MR ] × √
α1 [Q] is derived; the
√ O(1)
coefficient is not. TL = MR / 2π (z = MR /T = 2π) is
a convention, not the freeze-out of any computed process.
The degenerate heavy spectrum with real kernels has
three washout channels, K = (2.19, 8.38, 46.9); computing the relaxation with them, with finite-rate spectators,
the sphaleron conversion at the electroweak crossover and
a bias switched on at MR , multiplies the value above
by 0.272 (68%: 0.247–0.306; computed, with the plasma
inputs estimated). The print has no switch-on: its bias
never ends and runs from the start of the hot plasma, at
or above TL , and with the sphaleron rate of the printed
g2 = 2/3 the relaxation from that start is√computed below. With the Gaussian carrier Q = α/ π this would
give ηB = 1.66 × 10−10 , but that carrier is not realizable
(part (1)). The printed coupling fixes the object and its
charge. The transport couples only through the internal
µ
Dirac operator, as the current vertex A5µ JB−L
inside the
fermion term, which acts per particle and contains no
topological term. The line it transports is the determinant line of the chiral kernel. The right-handed neutrino
lies outside that kernel, since its Majorana mass is not
part of the printed DF (derived from the print, conditional), so cX = Ngen × (−1) = −3. The coupling acts
on B−L alone and requires only whole turns of the line

14
per transport period, so how far the line turns per flux
step is a stated input, and one turn of the whole line
per step would satisfy the coupling as well. The stated
carrier gives each generation a full B−L turn per step,
nw = 3 at one step per e-fold (Axiom AH.35): with the
determinant-line convention cX = −3, whose sign sets
the sign of the bias, |cX |/nw = 1 and Q = 2πα, the
per-e-fold deposit itself, counted once. Pinned to the
print, with the bias running from the reheating temperature, this gives ηB of about (2–3) × 10−9 , three to five
times Planck’s value, and no draw of the plasma inputs
falls below Planck’s value; started at MR the bias gives
1.9 × 10−9 , and started at TL , the lowest start the print
allows, 1.5 × 10−9 (computed). The sphaleron conversion
enters once, and the heavy tower of the microsector has
annihilated or decayed long before TL , so it does not
dilute the asymmetry (computed). The printed mechanism therefore overshoots. Agreement would need one
kernel winding per e-fold, which the stated carrier does
not give. The printed Heisenberg–Weyl clock of the three
kernel modes (Theorem AH.144) does not supply one. It
gives one kernel winding per flux step only if its labels
are read as charges, and the choice of window gives 0,
1 or 2; the printed flux step makes every multiplet flow
equally, which the clock’s zero spectral flow cannot reproduce; its unequal part biases B−L-conserving flavour
processes (both CKM links and six off-diagonal Dirac
Yukawas), which part (3) above excludes; and once the
three washout channels are resolved its value spreads over
4.5–7.3×10−10 , depending on a generation assignment the
print does not make (computed). Thermal leptogenesis
gives no asymmetry either, by the real-kernel protection
above (εi = 0). The baryon asymmetry is therefore carried as one calibrated input, ωb (App. FK, main volume),
and the printed transport’s first-principles value of ηB ,
about (2–3) × 10−9 , three to five times Planck’s value,
is recorded. A right-handed neutrino inside the kernel
would give no asymmetry at all.
Theorem AH.34 (Winding-form uniqueness). Assume
the single transport premise (E-transport): the internal
determinant line is transported by the cosmological expansion (its microsector modes are comoving). Then the
winding form of Theorem AH.33 is forced:

tion of the scale factor is a coordinate convention: under A → λA the holonomy must be unchanged for all
λ > 0, so G(λA) λ dA = G(A) dA, forcing G ∝ 1/A, i.e.
A5 ∝ d ln A; any other power d(An ) fails the scaling test,
and the Cauchy functional equation for a multiplicative-toadditive holonomy map gives the same conclusion. The coefficient Q is set in part (1) of the proof of Theorem AH.33:
its α1 power is derived, and its size rests on the stated
turn of the determinant line per flux step.
Status and caveats (no-fitting rule). (a) Winding
form: the form A5 = Q d ln A is not a bare insertion:
it is theorem-grade conditional on the transport premise
(E-transport) of Theorem AH.34 — the weakest residual premise (Postulate Y.10 as written supplies only a
static determinant-orthogonal offset; that the line is transported by the expansion remains an axiom-level input,
not a derived statement). The carrier is d ln A, the e-fold
count of the matter metric: the split of the scale factor between the arena factor and ψ̄ is pure gauge (main
volume, App. ZS), so there is no second carrier and no
factor β/p. The freeze-out point moves ηB by a factor
of order 3, and with the stated carrier the mechanism
gives about (2–3) × 10−9 , three to five times Planck’s
value (proof of Theorem AH.33, step (4)). (b) Sign:
sign(ηB ) = sign(Q) sign(H); H > 0 is C-even and sign(Q)
is an internal-orientation datum logically independent of
the external cosmic-time arrow, so the bare action derives
the α4 power of |ηB | but does not orient it. The orientation is supplied by Axiom AH.35 (adopted, priced, not
derived); all four candidate derivation routes for that orientation are closed by proof (see the axiom’s status note
below). Subject to (a)–(b), DFD derives the α4 power of
the baryon asymmetry, a quantity the Standard Model
misses by ∼ 8 orders, and its B−L bias; with the stated
carrier, its coefficient overshoots Planck’s value, giving
about (2–3) × 10−9 , three to five times Planck’s (proof
of Theorem AH.33, step (4)). Because A5 = Q d ln A is
exact (FA5 = 0), the step fires on the local e-fold count,
so the asymmetry follows the local expansion and carries
no isocurvature, SB = 0 exactly for an adiabatic onset of
the radiation era, and the baryons follow any adiabatic
primordial perturbation (computed); being locally pure
gauge, A5 acts only through B−L-violating processes.

A5 = Q d ln A
is the unique 1-form consistent with (i) reference invariance, (ii) segment additivity, and (iii) convention invariance of the scale-factor normalization; in particular
A50 = QH follows rather than being defined.
Proof. (i) Reference invariance. The holonomy must vanish on a static background: an explicit dt component
would require a preferred time unit and would wind in
a static universe — excluded. Hence A5 = G(A) dA
along the expansion flow. (ii) Segment additivity is automatic for 1-forms (the same additivity input as the
temporal-segment identification of App. Q, main volume). (iii) Convention invariance. The normaliza-

c.

The transport axiom and ωb

Axiom AH.35 (E-transport′ ). The E-transport advances
the S 3 flux by one lattice step per e-fold, oriented with
expansion.
The S 3 flux advanced here is not the sector label j of
the χ relic (main volume, App. AV): advancing that label
one step per e-fold would raise Ωχ h2 from 0.1193 to 0.126
after one step and to 0.41 by today (computed). This
clause is an adoption, not a derivation: we adopt it as
one priced discrete bit. All four candidate derivation
routes (lattice/degree, spectral, Chern–Weil, index) are

15
closed by proof : the unit rate is non-topological, rateinvisible to every operator-path functional, invisible to
the flat printed connection (Thm. AH.34’s A5 carries
FA5 = 0), and sign-blind under every index functional
(spectral flow is odd under loop reversal). Beyond its
coefficient (Theorem AH.33), the residue of the ωb chain
is therefore provably this one axiom-grade sentence; it is
minimal by construction, and it is priced: +1 discrete
bit in the theory’s accounting, which reads derived core
+ priced discrete bits (including E-transport′ ) + the four
continuous calibrations graded in App. FK of the main
volume (the vacuum coefficient R, the amplitude As , the
coefficient of ωb h2 , and the astrophysical τ shared with
ΛCDM), and the stated two-clock postulate (main volume,
App. ZS), which adds no parameter. With the stated
carrier, one flux step per e-fold turns each generation
once, and the mechanism gives about (2–3) × 10−9 , three
to five times Planck’s value, when every step is pinned
to the print (Theorem AH.33, step (4)). The axiom
was adopted after the conventional-point selection was
known, so it is priced as one discrete bit, never “zero
assumptions.” The refusal reading, declining the axiom,
leaves the deposit exponent w0 unfixed, so ωb keeps its
derived α4 power with its coefficient and its deposit factor
unfixed; it moves nothing else.
Theorem AH.36 (ωb at the conventional point under
E-transport′ ). Under Axiom AH.35 the deposit is quantized with the stated carrier: the unit rate gives |w0 | = 1,
and the expansion orientation gives sign(w0 ) = +1. The
per-e-fold deposit (1 + w0 2πα) is independent of the efold count N , so with w0√= +1, at the conventional
freeze-out
√ point TL = MR / 2π with the Gaussian carrier
Q = α/ π,


ηB = 0.2057 α4 1 + 2πα = 6.1007 × 10−10 ,
ωb h2 = 0.0222730,
against Planck 2018’s 0.02237 ± 0.00015. At full precision the base coefficient is 0.205705378693, i.e. ηB =
0.205705378693 α4 (1 + 2πα) = 6.1007 × 10−10 . This is the
conventional-point value that the CMB fits use (App. FK,
main volume). It is not the mechanism’s prediction: the
Gaussian carrier is not realizable, and with the stated
carrier the mechanism gives ηB of about (2–3) × 10−9
when every step is pinned to the print, three to five times
Planck’s value (proof of Theorem AH.33, step (4)).
What the axiom does not force. The axiom forces
w0 = +1 only. The identification N = 60 — the completion reading w0 N = kmax with the transport-torus
fiber index kmax = 60 (the fiber of the transported determinant line, not the χ relic’s sector label) — is dataconditional, not forced: the entropy-anchored e-fold count
is Nanchor = 57.76–59.37 across the three printed onset
conventions (thermal anchor 59.37), so the identification
sits 0.63–2.24 e-folds above them. The two values are
observationally degenerate in ηB at the 0.05% level (far
below the Planck 0.67% width), and ηB and ωb h2 above
are N -free, but the onset bookkeeping is a named open

item (main volume, open problems).
Grade. ωb h2 : the α4 power and the B−L bias current
are derived; the coefficient is not. Agreement with Planck
holds only at the conventional freeze-out point with the
Gaussian carrier, which is not realizable; with the stated
carrier, the mechanism gives ηB of about (2–3) × 10−9
when every step is pinned to the print, three to five times
Planck’s value, and agreement would need one kernel
winding per e-fold, where the stated carrier gives three.
ωb is therefore one calibrated input of the cosmological fits,
as in ΛCDM, with that value recorded as the mechanism’s
first-principles result. The chain uses the adopted, priced
Axiom AH.35; the axiom is one discrete bit, charged in
the knob ledger (App. FK, main volume) and in the trials
audit (App. CE row 9, main volume) exactly like every
other adopted bit.

d.

Transit Times and the Horizon

With the parent strain’s σ kept a collapsed object
has general relativity’s event horizon, at the areal radius
2GM/c2 (App. WS): radial transit times and the causal
structure outside it are Schwarzschild’s, and a signal
emitted inside it does not reach large radii, as in general
relativity.

e.

Quantum-Gravity Amplitude Finiteness

Theorem AH.37 (Finite All-Loop Amplitudes). The
L-loop graviton-graviton amplitude in DFD is a 604L term sum of bounded rational functions, bypassing the
Goroff–Sagnotti counterterm tower entirely.
Proof. Each loop integral in DFD’s QG sector is a trace
over Hmicro , finite-dimensional with dim = 60 (Theorem AH.13). For a four-graviton amplitude at L loops,
QL
the integration measure is i=1 TrHmicro(i) , a product of
L traces, giving 60L terms per leg insertion. With four
external legs, the total is 604L rational-function summands. Each summand is bounded because mode masses
Mn2 ≤ Λ2top = αMP2 are finite. The standard Goroff–
Sagnotti R3 counterterm at 2 loops in pure GR gravity
is absent here because the integrand is finite trace, not
a divergent loop momentum integral. Tree-level matches
GR through O(s2 /MP2 ).
Remark AH.38 (Status note). The finiteness argument
above rests on the identification of each loop integration
with a trace over the finite-dimensional microsector, which
is asserted at this site rather than derived here; bounded
summands alone would not give finiteness without that
identification and the finite term count it supplies.
Corollary AH.39 (No LHC Black Holes). Collisions at
LHC energies form no black holes.

16
Proof. With the parent strain’s σ kept, DFD’s gravity where µ → 1 is four-dimensional general relativity (App. WS), whose black-hole formation threshold
is set by the Planck energy; LHC centre-of-mass energies,
∼ 104 GeV, lie fifteen orders below it, and the truncated microsector has no Kaluza–Klein modes to lower it
(App. WS). Predicts null LHC mini-BH searches.

24. Embed AF into M60 (C) via the block-diagonal map:
C → e1,1 , H → M2 (C) ⊂ block of size 2, M3 (C) → block
of size 3. This uses 6 of the 60 dimensions; the remaining
54 encode the Berezin–Toeplitz tower modes that Connes’
construction does not natively contain. Hence AF ⊂
M60 (C) strictly.
i.

f.

With the parent strain’s σ kept, static stars obey
the TOV equations (main volume, App. AT, Theorem AT.13); for the reference Γ = 2, K = 100 polytrope
Mmax = 1.637 M⊙ , and PSR J0740+6620 (2.08±0.07 M⊙ )
is consistent with DFD as it is with GR.
g.

Primordial Gravitational Waves

Theorem AH.40 (Tensor-to-scalar ratio in the pre-inflation scenarios). DFD’s primordial tensor ratio is r ≲ 10−7
(main volume, App. SM). Tensor modes and the χ flux
coordinate share one cone and one mode equation, so any
stage that freezes tensors with power Pt also leaves χ a
variance Pt /(8α6 ) per e-fold, and the relic’s measured
abundance caps it. Of the two pre-inflation drivers, (a)
r = 0 satisfies the cap; (b), an external inflaton with
r = 16ϵ ≈ 4 × 10−3 (the value of Theorem AH.31), kicks
χ by 2.6 turns per e-fold and erases the relic, so it is
excluded on the main volume’s branch, which is DFD’s
cosmology. The Wheeler–DeWitt construction of §AH 12 b
gives 4α/π ≈ 9.3 × 10−3 on its own premises; that number
is the de Sitter dictionary’s 16ϵW , not the B-mode ratio.
Proof. By Theorem AH.29 no slow-roll-compatible scalar
exists on K = CP 2 × S 3 . Tensor modes hTT
ij during preinflation come from two possible drivers: (a) static initial
state with no inflation ⇒ r = 0; (b) external UV inflaton
with reheat at TRH ∼ 1015 GeV ⇒ r = 16ϵ, and ϵ ∼ 1/Ne2
−3
at Ne ≈ 60 gives r ≈ 4 × 10p
. With As = 2.1 × 10−9 ,
driver (b) needs H = π M̄P As r/2 ≈ 1.6 × 1013 GeV;
the massless χ then gains an rms of H/(2πfχ ) ≈ 2.6
turns per e-fold and is washed to the finite-space ceiling
Ωχ h2 = 0.139, excluded at +16σ. The main volume’s cap,
σθ2 < 1.9 × 10−3 over the ≃ 36 e-folds between today’s
horizon and χ’s oscillation onset, gives r ≲ 10−7 for any
power-law tensor tilt; (a) lies below it.
h.

Path Integral Quantization

Neutron Stars

Connes Spectral-Triple Connection

Proposition AH.41 (DFD/NCG Relationship). DFD’s
microsector algebra M60 (C) strictly contains Connes’
AF = C ⊕ H ⊕ M3 (C) as a sub-structure carrying the
Berezin–Toeplitz tower up to kmax = 60.
Proof. Connes’ finite spectral triple (AF , HF , DF ) has
algebra AF = C ⊕ H ⊕ M3 (C), real dimension 2 + 4 + 18 =

Theorem AH.42 (Factorized DFD Measure). The full
DFD measure factorizes as a coarse-grained 4D continuum
measure times an exact Lebesgue measure on C60 .
Proof. The Hilbert space factorizes as HDFD = F ψ ⊗
Hmicro with dim Hmicro = 60. The path-integral measure
is Dψ · D60 ζ where ζ are the microsector amplitudes. The
microsector measure is the standard Lebesgue measure on
C60 , finite and non-singular. The continuum Dψ measure
is regulated by the same finite microsector cutoff Λtop ,
so the entire QFT measure is well-defined modulo the
standard 4D continuum issues. No regulator divergence
appears at any loop order from the microsector trace.
Theorem AH.43 (Ghost-Free Spectrum). Below the
spectral cutoff the propagator has a single positive-residue
pole at k 2 = 0 (above it the Weyl-squared term adds the
spin-2 pole of App. WS).
Proof. Under Axiom WS.1 the scalar ψ is an instantaneous constraint with no propagating pole. For hTT
ij : by
Theorem AH.20 the propagator is block-diagonal and the
tensor block has a positive-residue pole at k 2 = 0.

j.

Topology Change of Optical Slices (Optional)

Theorem AH.44 (Topology-Change Admissibility).
DFD admits smooth topology change of its optical slices,
the level surfaces of ψ, because its causal ordering is guaranteed by the absolute time t of the fixed flat arena, which
every characteristic advances, and the optical metric is
constructed, not a fundamental spacetime metric; space
itself, the arena R3 , keeps its topology.
Proof. Geroch’s theorem (1967) states that smooth topology change is impossible in 4D Lorentzian manifolds
without a closed timelike curve, where causality is the
fundamental structure. In DFD, the spatial slice is
fixed flat R3 at the fundamental level; the optical metric
g̃ = −c2 e−2ψ dt2 + dx2 is constructed from ψ. Topology
change of the optical slices (level surfaces of ψ) is admissible because the fundamental metric is not changing.
Geroch’s theorem applies to the unchanging fundamental
flat R3 , vacuously.

17
5.

Frontier Territories: Particle and Astrophysical
Predictions

Seven experimental fronts follow.

a.

Muon and Electron g − 2: Structurally Negligible

b.

QCD Confinement and Glueball Spectrum

Theorem AH.47 (Confinement: area-law form open).
The area-law form is not derived in the proof below. The
Wilson loop in DFD-QCD obeys an area law


⟨W (C)⟩ ∼ exp − σDFD Area(C) ,
1/2

σDFD = 440 ± 25 MeV
Theorem AH.45 (DFD Anomalous Magnetic Moment
Suppression). The DFD-specific contribution to the lepton
anomalous magnetic moment is bounded by

2
1 mℓ
δaDFD
≤
,
ℓ
α Λtop
√
where Λtop = α MP . Numerically, δaDFD
≤ 1.4 × 10−36
µ
DFD
−41
and δae
≤ 3.3 × 10 , about 26 and 28 orders of magnitude below FNAL and Hanneke sensitivity respectively.
Proof. Heavy-mass decoupling (Appelquist–Carrazone).
DFD’s only beyond-QED degrees of freedom are the microsector modes with mass ∼ Λtop . The leading vertex
correction from a heavy mode of mass M to a low-energy
observable scales as (mℓ /M )2 relative to the leading QED
contribution α/(2π). Hence δaℓ ∼ (α/2π)(mℓ /Λtop )2 ·
(1/α) = (mℓ /Λtop )2 /(2π), conservatively
bounded by
√
(mℓ /Λtop )2 /α. With Λtop = αMP = 1.04 × 1018 GeV:
(mµ /Λtop )2 = 1.04 × 10−38 , multiplied by 1/α: δaµ ≈
1.4 × 10−36 . For electron (me = 5.11 × 10−4 GeV):
(me /Λtop )2 = 2.4 × 10−43 , δae ≈ 3.3 × 10−41 .
Corollary AH.46 (FNAL gµ − 2 Tension is Not a
DFD Effect). The historic ∼ 4–5σ FNAL tension (at
∆aexp−SM
≈ 245 × 10−11 against the 2020 data-driven
µ
SM) is 27 orders of magnitude larger than DFD’s bound
and therefore cannot originate in DFD. As of the 2025
Muon g − 2 Theory Initiative White Paper — with the
BMW/lattice HVP consensus and the CMD-3 e+ e− data
shifting the SM prediction upward by ∼ +223 × 10−11 —
this tension has collapsed to ∼ 0.6σ: the anomaly is now
largely removed. This supports DFD’s structural no-newmuon-physics prediction (Theorem AH.45), which forces
δaDFD
≤ 1.4 × 10−36 with no free parameter and hence
µ
SM
aµ = aµ .
Proof. Direct comparison against the historic discrepancy:
245×10−11 /1.4×10−36 = 1.75×1027 . By Theorem AH.45
the DFD contribution is structurally bounded by the decoupling argument and cannot grow without invalidating microsector finiteness; DFD therefore commits to
aµ = aSM
and to any residual tension resolving outside
µ
DFD (HVP refinement, lattice, or systematics). The 2025
consensus did exactly this — the SM hadronic-vacuumpolarization contribution rose, driving the tension to sub1σ — so the structural null is now favored by the data.
No DFD-specific central value is introduced; the sole DFD
claim is the decoupling upper bound.

(normalization lattice-calibrated).
Proof. The Wilson loop is computed as a trace over
Hmicro (finite-dimensional,
dim
= 60), ⟨W (C)⟩ =


H
(1/60)TrHmicro [P exp C iA(3) ]. The string tension is
fixed by the spectral spread of the gauge-charge operator on Hmicro , computed via Casimir invariants:
p
1/2
σDFD = (ΛQCD /2π) kmax /Ngen · Fgeom with ΛQCD =
MP α19/2 = 61.20 MeV, kmax /Ngen = 60/3 = 20,
Fgeom ≈ 1 (Fubini–Study volume normalization). Numerical status:pthe derivation-grade content is the scaling
σ 1/2 ∝ ΛQCD kmax /Ngen ; the printed coefficient chain
does not close to 440 MeV (it evaluates to
p ≈ 268 MeV
at Nf = 3 if one includes an undefined 1/ 1/Nf factor,
≈ 154 MeV without it — see the status note below), so the
absolute normalization is lattice-calibrated to 440±25 MeV
(main volume, App. AQ: ≈ 465 MeV at r0 = 0.5 fm), not
forward-derived.
Theorem AH.48 (Lightest Glueball 0++ Mass). The
DFD target m0++ = 1.69 ± 0.10 GeV is tested by the
main volume’s App. AQ run (1.71 ± 0.17 GeV) and is
consistent with f0 (1710); its normalization is open (status
note below).
Proof. The lightest glueball corresponds to the lowest
non-trivial eigenvalue of the gauge-only Laplacian on
CP 2 in the singlet 0++ representation. The Fubini–
Study eigenvalue spectrum on CP 2 has the singlet ladder
2
2
λn = n(n + dimC CP 2 )/rFS
= n(n + 2)/rFS
, with rFS
the Fubini–Study radius. The lowest non-trivial eigen2
value is at n = 1: λ1 = 3/rFS
. Translating to mass
2
2
via m = λ1 · ΛQCD, 3−flavor · (4π) (MS matching factor):
p
m0++ = 8ggeom · ΛQCD,3−fl with ggeom = 3/(2π)2 from
the Fubini–Study spectral
density. With ΛQCD,3−fl =
√
210 MeV: m0++ = 8 · 0.076 · 210 MeV = 1.69 GeV.
Threshold corrections from the η ′ -anomaly sector add
δthr ≈ ±0.10 GeV.
a. Status note (normalization and flavor scheme open).
As
p printed, the numerical chain above does not close:
8ggeom Λ = 0.78 × 210 MeV ≈ 164 MeV, while the in√
termediate form m2 = λ1 Λ2 (4π) gives 12π ×210 MeV ≈
1.29 GeV — the two printed coefficient forms are mutually
inconsistent and neither reproduces the stated 1.69 GeV.
Separately, the input labeled ΛQCD,3-fl =
√ 210 MeV is
numerically the five-flavor MS value (cf. 4π × 61.20 =
217 MeV in this same chain); the three-flavor value derived from the identical MP α19/2 chain in the τ -sector
update is ΛQCD,3 = 332 ± 20 MeV. What holds at

18
derivation grade is the scaling m0++ ∝ ΛQCD from the
Fubini–Study eigenvalue argument; the overall normalization constant and the flavor-scheme assignment are
open items. The ledger value m0++ = 1.69 ± 0.10 GeV is
retained as the DFD target — consistent with quenched
lattice spectroscopy (the Appendix AQ companion run
of the main paper gives 1.71 ± 0.17 GeV) — but its normalization from the CP 2 spectral argument is not yet
derivation-grade. The same caveat applies to the adjacent
p AH.47): as printed,
√ string-tension chain (Theorem
√
20/(2π) · 61.20 MeV · 4π/ 1/Nf evaluates to ≈
268 MeV at Nf = 3, not 440 MeV; the scaling structure stands, while the printed normalization factors are
open items.

c.

Electroweak Phase Transition and Sphaleron Rate

Theorem AH.49 (EWPT Crossover and Critical Temperature). TCDFD = 159.5 ± 1.5 GeV (crossover), with the
measured mH /mW = 1.56 > 0.9 forcing crossover.
Proof. DFD’s Higgs sector is identical to SM at the electroweak scale (Theorem AH.17 gives the exponent of v =
246.09 GeV, and the conjectured λH = 1/8 gives mH ≃
124.6 GeV at the top-quark scale (main volume, App. Z);
the argument below uses the measured mH ). Hence the
finite-T effective potential V (ϕ, T ) is the SM expression.
Lattice SM at mH = 125 GeV (D’Onofrio–Rummukainen
2016) gives crossover at TC = 159.5 ± 1.5 GeV. The firstorder line ends at mH ≃ 72 GeV (mH /mW ≃ 0.9; Rummukainen, Tsypin, Kajantie, Laine and Shaposhnikov
1998), which the measured mH /mW = 1.56 exceeds.
Theorem AH.50 (Sphaleron Rate at Reheating).
Γsph /V ≈ 4.8 × 1046 GeV4 at TRH = 2.4 × 1013 GeV,
a reheating temperature that no printed mechanism supplies (main volume, App. Q).
5
Proof. Above EWPT, Γsph /V = καW
T 4 with lattice
κ = 18 ± 3 (D’Onofrio–Rummukainen–Tranberg 2014).
The printed g2 = 2/3 runs to αW = 1/41.5 at TRH
(computed), so Γsph /V = 18 · (1/41.5)5 · (2.4 × 1013 )4 ≈
4.8 × 1046 GeV4 . The per-particle rate is (Γsph /V )/T 3 =
3.5 × 106 GeV, and the B+L relaxation rate carries a multiplicity between 13Nf /4 ≃ 10, with every Yukawa coupling in equilibrium, and 6Nf = 18, with none (Burnier,
Laine and Shaposhnikov, hep-ph/0511246): ΓB = (3.4–
6.3) × 107 GeV. Against the radiation-era Hubble rate
√
H = 1.66 g∗ T 2 /MP = 8.1 × 108 GeV (g∗ = 106.75),
ΓB /H ≈ 0.04–0.08 (≈ 0.14–0.27 in the convention
H = T 2 /M̄P ): the sphalerons are out of equilibrium
at TRH . Since ΓB /H grows as 1/T , they come into equilibrium at 1–2 × 1012 GeV (3.5–6.4 × 1012 GeV in the
second convention; computed), far above the electroweak
crossover, which is all the conversion factor uses.

Corollary AH.51 (Baryogenesis efficiency: α4 power
and B−L bias derived). Thermal leptogenesis gives no

asymmetry: the minimal real-kernel lepton sector gives
Im[(Yν† Yν )2ij ] = 0, hence εi = 0 and the thermalminimal DFD
leptogenesis ηB
= 0 (status note, §AH 4 b).
The canonical DFD position is the internal axial Berryholonomy channel (Theorem AH.33): DFD takes the
identified scale MR = 4.744 × 1012 GeV and forces the α4
power of the asymmetry, and its bias current is B−L. At
the conventional freeze-out point with the Gaussian carrier
base
the base value is |ηB
| ≃ 0.206 α4 ≃ 5.8 × 10−10 , completed by the transport deposit to ηB = 0.206 α4 (1+2πα) =
6.1007 × 10−10 (Thm. AH.36); that carrier is not realizable, and with the stated carrier, pinned to the print,
the mechanism gives about (2–3) × 10−9 , three to five
times Planck’s value (Theorem AH.33). The sign (matter vs. antimatter) is fixed by the expansion orientation
of Axiom AH.35, a CP-odd datum carried by A5 on the
determinant line — adopted, not derived. (The strongwashout efficiency κf = 4.87 × 10−3 quoted in the status
note above belongs to the thermal chain and does not
enter the Berry channel. The washout and reprocessing
of the Berry channel itself, with three washout channels,
finite-rate spectators and the relaxation computed, multiply the conventional-point value by 0.272; proof of Theorem AH.33, step (4).) Baryogenesis is therefore a DFD
result whose α4 power and B−L bias are derived and
whose printed mechanism gives about (2–3) × 10−9 , three
to five times Planck’s value, with the sign fixed by an
adopted axiom; it is not a closed prediction.

d.

Multimessenger Binary Neutron Star Signatures

With the parent strain’s σ kept a binary neutron-star
merger is general relativity’s where µ → 1 (App. WS;
argued for fully nonlinear fields): the post-merger peak
frequency and the remnant’s ringdown are GR’s, and
there is no echo train.

e.

Proton Lifetime

Theorem AH.52 (Proton Stability). If baryon number is
violated only by the electroweak anomaly, Γ(p → e+ π 0 ) =
0 and Γ(p → ν̄K + ) = 0 exactly.
Proof. Sphaleron transitions change B and L together,
by ∆B = ∆L = ±Ngen = ±3 (from the chiral anomaly
with three SM generations). Standard p → e+ π 0 has
∆B = −1, not a multiple of 3, hence forbidden. Similarly
p → ν̄K + is forbidden. The minimal allowed processes
change baryon number by three units, ∆B = ∆L = ±3
(main volume, App. F).
Lemma AH.53 (Allowed-Channel Lifetime). The
∆B = ±3 rate is suppressed by exp(−Sinst /ℏ) =
exp(−4π/αW ) ≈ exp(−373).
Proof. Electroweak instanton action is Sinst = 4π/αW ≈
4π · 30 ≈ 373 at the EW scale. At T = 0, the tunneling

19
rate is Γ ∼ mp · exp(−Sinst ). Hence τ = exp(373)/mp ≈
10130 yr. Even with the ∼ 1035 nucleons of Hyper-K and
1010 years, expected events ∼ 1035 · 1010 /10130 = 10−85 –
unobservable.
Corollary AH.54 (Hyper-K, DUNE, JUNO Null Prediction). If baryon number is violated only by the electroweak
anomaly, DFD predicts zero signal at every projected sensitivity in standard channels.
Proof. Hyper-K targets τp > 1035 yr on p → e+ π 0 and
p → ν̄K + . By Theorem AH.52 both are forbidden. By
Lemma above, the allowed multi-baryon channel has τ >
10120 yr. Therefore expected events at any current or
planned proton-decay experiment is exactly zero from
DFD. A single positive event in any standard channel
falsifies the selection rule of Theorem AH.52.
f.

Holographic Entropy Bounds

Theorem AH.55 (Bulk Information Density). ρinfo =
log2 (dim Hmicro )/ℓ3P = log2 (60)/ℓ3P ≈ 5.91 bits per
Planck volume.
Proof. Each Planck-volume cell carries Hmicro as its quantum state space. Maximum information storable in a cell
of Hilbert dimension d is log2 d bits (standard quantum
information). With d = 60: log2 60 = 5.91 bits per cell.
Density: 5.91/ℓ3P bits/volume. Note this is finite because
dim Hmicro < ∞, in contrast with continuum QFT where
the per-volume capacity formally diverges.
Corollary AH.56 (Λ-Topology Reciprocity). SdS ·
Λ/MP4 = O(1).
Proof. By Theorem AH.23 and the clock dictionary, the
observed dark-energy density is Λ/MP4 = ΩΛ (3/8π)Cα57
(main volume, App. AU). By Theorem AH.22, H02 =
Cc5 α57 /(Gℏ). The de Sitter entropy from Gibbons–
Hawking is therefore SdS = (πc5 )/(GH02 ℏ) = π/(Cα57 ).
The clock coefficient C cancels in the product: SdS ·
Λ/MP4 = 3ΩΛ /8 = O(1).
g.

Ultra-High-Energy Cosmic Rays and Photon Dispersion

Theorem AH.57 (DFD GZK Threshold = SM).
DFD
Eth
= 1.07 × 1020 eV, identical to SM at the 10−14
level.
Proof. The GZK threshold satisfies Ep · εCMB = mπ (mp +
mπ /2) at the ∆+ resonance. Pion mass mπ and proton mass mp in DFD descend from the same spectral
construction as α−1 and the fermion mass ladder (Theorem AH.27); their numerical values match PDG to ≤ 1%.
The CMB photon distribution is unchanged by DFD
(the cosmological photon sector is not modified). Hence
DFD
Eth
equals SM. Any structural correction is at the
Padé-protected O((E/EPl )6 ) order, giving relative shift
(1020 eV/1028 eV)6 = 10−48 – negligible.

Theorem AH.58 (No Lorentz-Violating Photon Dispersion). The DFD LIV parameter ξLIV ≡ 0 identically.
Proof. DFD’s optical metric is conformally flat on equipotential slices (ψ = constant): g̃ = −c2 e−2ψ dt2 + dx2
has zero Weyl curvature in the spatial sector. The dispersion relation on such a slice is ω 2 = c2 |k|2 exactly,
with no ξ(p2 /MPl )n correction (those would require nonconformal slice metric). Off-slice variations of ψ produce
only conformal-factor changes, which preserve the lightcone structure. By Theorem AH.20, no derivative coupling between ψ and hTT exists; in particular, no quartic
LIV operator. Hence ξLIV = 0 structurally. Pierre Auger
bounds |ξ| < 10−21 are consistent by structural identity.
Future tightening cannot “test DFD harder” on LIV –
the prediction is rigid zero.

6.

Standard Model Lagrangian, BSM Constraints,
and Cosmology

The Topological Pre-Inflation construction is kept as a
conjectured extension beyond DFD’s stated scope, not a
DFD prediction (Sec. AH 6 b).

a.

Vacuum Stability

An independent recomputation graded the vacuumstability theorem of Sec. AH 2 a questionable; the disagreement is a units artifact, not a physics conflict.
Theorem AH.59 (Vacuum Stability in the Dimensionless Variable). In the dimensionless ratio δV /MP4 , one(1)
loop vacuum feedback satisfies |δVCW /MP4 | ≤ O(α2 ). The
113
Sec. IX.H “λ ∼ 10
violently unstable” warning is a
presentation defect: it implicitly uses ρc (a cosmological output) as if it were the ψ effective-mass input. The
Planck-scale quantity in that denominator is the Newtonian gradient stiffness c4 /(32πG) ∼ MP2 of the ψ kinetic
term, not ρc ; ψ itself carries no potential term (Theorem AH.136, sub-(i)).
(1)

Proof. Theorems AH.1 and AH.3 establish |δVCW /MP4 | ≤
O(α2 ) in the dimensionless trace over Hmicro . The 10113 of
the main text is obtained by inserting ρc ∼ H02 MP2 /(8πG)
into the ψ self-energy denominator, but ρc is itself an
output of the exponent identity GℏH02 /c5 = Cα57 (Theorem AH.23; C measured), not a fundamental input. The
correct denominator is the Planck-scale Newtonian gradient stiffness of the ψ kinetic term, c4 /(32πG) ∼ MP2
(“ψ effective mass” is a mislabel: ψ has no potential
term, Theorem AH.136, sub-(i)). Substituting MP2 instead of ρc removes the 10113 enhancement and recovers
the dimensionless α2 bound.
Remark AH.60 (Verification Reconciliation). One computation works in the dimensionless variable (correct),
the other in the dimensional ρvac variable (also correct

20
for that variable); they agree on the underlying integrals.
The presentation defect is the carrying of the dimensional
variable, with ρc in the denominator, into Sec. IX.H; that
section states the dimensionless-variable bound beside the
estimate, and this appendix keeps both readings.

b.

Topological Pre-Inflation Status

The TPI construction (Theorems AH.29–AH.31) is not
derivable from DFD’s stated structure (five independent
checks; Remark AH.109), and Theorem AH.136 excludes
a slow-roll-compatible native inflaton. The TPI material
is kept as a conjectured extension beyond DFD’s stated
scope; it is not a DFD prediction.
Remark AH.61 (Status note: the TPI staircase). The
integers {1, 12, 30} appearing in the TPI staircase
do not match DFD’s canonical cohomological catalog
{3, 7, 8, 13, 19, 31, 49, 57, 60, 108, 137}. The Lichnerowicz
rigidity argument excludes graviton-mode inflatons but
does not exclude an external scalar inflaton coupled to
DFD; therefore the “no slow-roll” premise is narrower
than the stated theorem. Independent recomputation
of ns , As , r without the construction’s internal staircase
parameters does not converge to Planck values. The
verification record is in Remark AH.109.

c.

CKM Per-Channel Agreement

The magnitude integers are suggestive of cohomological
origin, and the CP-even apex is fixed at ρ̄ = 43
2 α = 21.5α
by the Euler-projection postulate.
Remark AH.62 (Per-channel CKM check). Three independent recomputations against PDG 2024 (λ =
0.22501 ± 0.00068, A = 0.826+0.016
−0.015 , ρ̄ = 0.1591 ± 0.0094,
η̄ = 0.3523+0.0073
)
give
per-channel
deviations λ at
−0.0071
∼ 0.5%, η̄ at ∼ 1.5%, A at ∼ 4.6%, and ρ̄ at ∼ 1.4%
for the Euler-projected apex ρ̄ = 43
2 α (the raw assignment ρ̄ = 19α gives a ∼ 13% deviation). The pattern
(31, 108, 43
2 , 49) × α has no continuous fitted parameter;
its integers are selected (App. FK of the main volume)
and the apex is fixed by the Euler-projection postulate
(App. AO).

d.

Full Standard Model Lagrangian from Microsector

Theorem AH.63 (SM Lagrangian Derivation). Every
coefficient of LSM = Lgauge + Lfermion + LHiggs + LYukawa +
Lθ + LGF + Lghost + LνR is given a structural source
on CP 2 × S 3 at kmax = 60 or is counted among the
inputs of App. FK of the main volume (the fitted Higgs
prefactor, the Higgs quartic, the imported ηq and the
discrete selections); the absolute charged-fermion Yukawas
additionally carry data-selected labels, graded separately

in App. FM of the main volume, and the statement here
is not a derivation of that absolute spectrum.
Proof. Gauge sector. The (3,2,1) partition + singlet on the bundle E = O(9) ⊕ O⊕5 carries the gauge
group SU (3) × SU (2) × U (1) as a dictionary entry: the
electroweak factor is the isotropy group of CP 2 under
one stated identification, and colour is a stated input
(main volume, App. F; Theorem AH.9). Couplings:
sin2 θW = 3/13 (App.√Z), and αs (MZ ) = 0.1187 from the
stated strong scale 4π MP α19/2 (main volume, open
problems).
Fermion sector. Three generations are a stated input
identified with χtop (CP 2 ) = 3, and the representations
of a family are inputs, as in the Standard Model; their
hypercharge twists use the derived flux quantum q1 = 3
(main volume, App. F).
√
Higgs sector. v = MP α8 2π = 246.09 GeV, with
the exponent derived and the prefactor fitted (Theorem AH.17); the conjectured λH = 1/8 from dimension
counting (main volume, App. Z).
√
Yukawa sector. yf = Af αnf / 2 via the Berrybundle overlap integral on CP 2 (Theorem AH.27). Each
prefactor Af is a product of four blocks (kernel, QCD,
Dirac, generation); for seven of the nine charged fermions
the absolute prefactor is data-selected (main volume,
App. FM).
θ-term. θQCD = 0 at tree level, loop-bounded, with
θbare = 0 fixed by the spectral action given that nothing is
added to it for the gauge fields (Theorem AH.66 below).
Gauge-fixing + ghost. Standard BRST quantization
on the 60-dim Hilbert space; finite-dim ghost sector arises
from the same Toeplitz truncation.
Right-handed neutrinos. MR = MP α3 = 4.744 ×
12
10 GeV (App. P; machine-verified to 40 digits).
P
cancellation.
Gauge anomalies i Yi3 = 0,
PAnomaly
P
a b
i Yi Ti Ti = 0,
i Yi = 0 vanish for the Standard Model
content of one family, which is a stated input; the flux
quantum does not force them.
Custodial SU (2). ρ = m2W /(m2Z cos2 θW ) = 1 at tree
level by the Higgs being a doublet under SU (2)L ⊗SU (2)R ;
structurally derived from the (3,2,1) partition.
e.

Beyond-Standard-Model Constraints

DFD’s microsector finiteness imposes sharp BSM exclusions.
Theorem AH.64 (BSM Forbidden List). The following BSM scenarios are forbidden at theorem level, items
(iv)–(vi) given the adopted gauge group (its electroweak
factor rests on one stated identification and colour is an
input; main volume, App. F): (i) supersymmetry, (ii)
QCD axion/Peccei–Quinn, (iii) magnetic monopoles, (iv)
dark photon U (1)′ , (v) leptoquarks, (vi) Z ′ bosons, (vii)
cosmic strings, (viii) domain walls, (ix) composite Higgs,
(x) extra macroscopic spatial dimensions, (xi) Lorentzviolating photon dispersion. There is no postulated dark

21
matter and no thermal WIMP candidate; the cold dark
matter is instead the derived χ-matter field (App. AV),
the harmonic b3 three-form on the internal S 3 = SU (2)
— a topology-forced geometric mode, not an added particle and not a Peccei–Quinn axion (item ii). DFD adds
no free dark-sector parameter: χ supplies the cosmological clustering scaffolding while the µ-law (µ(x), a0 ;
Theorem AH.21) covers galactic-scale phenomenology; inside galaxies the two compose by the adopted floor law
with a stated three-dimensional filling rule (main volume,
App. AV).
Proof. Each scenario fails a specific DFD axiom or theorem.
(i) SUSY. Adding even one superpartner forces kmax ≥
65 to accommodate the additional internal modes: the
closed Spinc index χ(CP 2 , E) = χ(O(9)) + 5 χ(O) =
55 + 5 = 60 is already saturated (equivalently, 60 = 3
geometric zero modes + 57 nonzero KK modes), so minimal padding admits no further internal mode (Theorem AH.9). A hypercharge-charged superpartner additionally violates the hypercharge-twist integrality condition (the q1 Y + 32 ∈ 12 Z requirement on Lq1 Y , App. F); a
would-be gauge-singlet (Y = 0) superpartner direction,
to which the twist integrality has no content, is instead
excluded by geometric rigidity — the Koiso/Lichnerowicz
zero-mode classification with b1 (K) = 0 leaves only the
trace ψ, the truncated microsector carries no size or shape
modulus, and the vacuum moduli space has dimension 0
and cardinality 1 (Theorem AH.179).
(ii) Axion. Theorem AH.66 below: θQCD = 0 at tree
level (θbare = 0 fixed by the spectral action given that
nothing is added to it for the gauge fields), loop-bounded
below 10−16 . Peccei–Quinn relaxation is unnecessary,
and a new Peccei–Quinn axion — an added U (1)PQ field
introduced to solve strong-CP — would correspond to
a 60th-mode-violating internal direction and is forbidden. This does not forbid the χ-matter field of App. AV:
χ is the pre-existing harmonic b3 three-form on S 3 (a
bosonic geometric zero-mode, not one of the 60 chiral
Spinc modes), it is not introduced to solve strong-CP,
and it carries no Peccei–Quinn charge. It is categorically
distinct from the forbidden PQ axion.
(iii) Monopoles. π2 (S 3 ) = 0 and π2 (CP 2 ) = Z
structure does not admit topological monopole solutions
in DFD’s (3,2,1) gauge embedding because the second
homotopy group is exhausted by the Kähler class.
(iv) Dark photon. The gauge group is SU (3) ×
SU (2) × U (1) from (3,2,1)+singlet, adopted rather than
forced: its electroweak factor rests on one stated identification and colour is an input (main volume, App. F).
Adding U (1)′ would require a sixth chiral multiplet, ruled
out by minimal-padding theorem (Theorem AH.9).
(v) Leptoquarks. Mediating quark-lepton transitions
requires gauge bosons charged under both SU (3) and
SU (2)L . No such bosons exist in E.
(vi) Z ′ . Same argument as (iv).
(vii) Cosmic strings. The electroweak vacuum manifold (SU (2) × U (1)Y )/U (1)em ≃ S 3 has π1 = 0. No

topologically stable string solutions.
(viii) Domain walls. The ψ-vacuum is unique by the
strict monotonicity of µ(x) = x/(1 + x) on [0, ∞).
(ix) Composite Higgs. The Higgs is the unique
(1, 2, +1/2) zero mode of the Spinc Dirac operator on CP 2
in E (App. H). Compositeness would require a stronglyinteracting subsector, absent from LSM .
(x) Extra macroscopic dimensions. The 4D spacetime is fixed flat R3,1 ; the additional 7 internal dimensions
are bundle directions of CP 2 × S 3 at fixed compact size,
not large macroscopic dimensions.
(xi) LIV. Theorem AH.58: ξLIV ≡ 0.
Corollary AH.65 (Sharpest BSM Null Predictions).
The most experimentally-decisive null predictions are:
(a) Detection of any squark/gluino/neutralino at LHC,
FCC-hh, or any future collider ⇒ DFD falsified.
(b) Observation of p → e+ π 0 or p → ν̄K + at any rate
at Hyper-K, DUNE, or JUNO ⇒ DFD falsified.
(c) Detection of QCD axion at any haloscope (ADMX,
IAXO, ABRACADABRA, CASPEr, MADMAX)
across the KSVZ/DFSZ window ⇒ DFD falsified.
(d) Detection of a millicharged particle — any charge
εe, any mass, any production channel (accelerator,
cosmic-ray, or terrestrial accumulation) ⇒ DFD
falsified.
Proof. Each item is a direct consequence of Theorem AH.64. (a) SUSY’s superpartners would force
kmax ≥ 65, exceeding the saturated minimal-padding
index χ(CP 2 , E) = 60 (Theorem AH.14 fixes kmax = 60);
hypercharge-charged superpartners additionally violate
the hypercharge-twist integrality condition (App. F), and
any gauge-singlet superpartner direction is excluded by
the single-vacuum geometric rigidity (Theorem AH.179).
(b) Standard proton-decay channels are forbidden by Theorem AH.52; a positive observation in any standard channel
violates its selection rule. (c) The QCD axion’s existence
requires the Peccei–Quinn relaxation mechanism, which
is unnecessary in DFD because θ̄ = 0 at tree level, with
θbare = 0 fixed by the spectral action given that nothing is
added to it for the gauge fields, and loop-bounded by Theorem AH.66; a haloscope detection in the KSVZ/DFSZ
window would imply a non-vanishing axion-photon coupling that DFD’s spectral-pairing closure forbids. (d)
follows from item (iv) of Theorem AH.64: a kineticmixing millicharge requires the forbidden dark-photon
U (1)′ , and a bare millicharge is a new charged internal
mode excluded by the saturated index χ(CP 2 , E) = 60;
accumulator-augmented shielded-Cavendish tests (Berlin,
Bogorad, Graham, and Ramani 2026; arXiv:2510.25825,
2510.25834) reach the irreducible cosmic-ray-produced
population within years, so the null is dated as well as
one-sided. Each of (a)–(d) corresponds to a single theorem
failing.

22
f.

Strong-CP Closure: Tree Level, Loop-Bounded

Theorem AH.66 (θQCD = 0 at Tree Level,
Loop-Bounded). With θbare = 0 fixed by the spectral
action given that nothing is added to it for the gauge
fields, the strong-CP angle θ̄ = 0 at tree level, with
loop corrections at the Standard-Model level (≲ 10−16 ;
App. AO). The neutron electric dipole moment from
this θ̄ sector is therefore at most ∼ 2 × 10−32 e·cm
(dn ≃ 2.4 × 10−16 θ̄ e·cm), comparable to the CKM background; the residual EDM is set by the CKM phase alone
and is dominated by the long-distance chiral pion-loop,
dDFD
∼ (0.9–5.6)×10−32 e·cm (Corollary below; App. AT
n
neutron-EDM ledger).
Proof. The CP mapping torus on X = CP 2 ×S 3 is TCP =
(X ×[0, 1])/(x, 0) ∼ (CP (x), 1), with dim TCP = 7+1 = 8
(even). On any closed even-dimensional oriented manifold
carrying a Z2 -graded Clifford-module bundle (here W =
S ⊗ E ⊕ S̄ ⊗ Ē with the antilinear CP lift; TCP itself
admits no Spinc structure, App. L), the chirality operator
Γ anticommutes with the Dirac operator: {Γ, DTCP } = 0.
Therefore the spectrum is ±-paired: every eigenvalue λ
has matching −λ. The Atiyah–Patodi–Singer η-invariant
X
η(DTCP ) = lim
sgn(λ) |λ|−s
s→0+

θ̄ ≲ 10−16 . The dominant neutron-EDM contribution is
the long-distance chiral charged-pion loop (Khriplovich–
Zhitnitsky; Seng, Phys. Rev. C 91, 025502 (2015)),
dSM,LD
∼ (1–6) × 10−32 e·cm, rescaled by the DFD
n
∼ (0.9–5.6) ×
Jarlskog ratio JDFD /JSM = 0.94 to dDFD
n
10−32 e·cm. The short-distance three-loop quark/chromo−34
electric channel (dSD
e·cm) is subdominant by
n ∼ 10
2–3 orders and is not the only contribution; the residual
θ̄-channel is ≲ 2 × 10−32 e·cm for the upper-end induced
θ̄ ∼ 10−16 , comparable to the CKM background. (A
short-distance-only estimate, ∼ 10−37 e·cm, omits the
dominant long-distance channel; the value here is set by
the App. AT neutron-EDM corollary.) The electron EDM
sits at the tiny CKM/SM floor, de ≲ 10−38 e·cm; the
leptogenesis Majorana channel is inert, because DFD’s
leptonic CP-odd invariant Im[(Yν† Yν )2 ] = 0 (the same
real-kernel structure that gives the ε1 = 0 baryogenesis
no-go), leaving only the CKM contribution.

g.

Precision QED Beyond g − 2

Other precision-QED channels (Lamb shift, positronium, muonic-H, helium HFS) receive δ ∼ (mℓ /Λtop )2 ∼
10−43 –10−38 DFD corrections, structurally invisible.

λ

vanishes term-by-term by spectral pairing for Re s > 8,
where the series converges absolutely by Weyl’s law, and
hence identically after continuation to s = 0. The Dai–
Freed factor entering the partition function is
ACP = exp(iπη(DTCP )/2) = 1
exactly, hence θ̄CP
=
0.
Combined with
arg det(Mu Md ) = 0 at tree level for det-orthogonal
mass matrices positive on the branch of Postulate Y.10
(App. AO) and the bare angle θbare = 0, fixed by the
spectral action given that nothing is added to it for
the gauge fields (main volume, App. L), this enforces
θ̄ = 0 at tree level. Loop-by-loop ledger: at tree level,
θtree = 0; the shift of arg det(Mu Md ) vanishes identically
through Yukawa degree eleven for Hermitian tree-level
mass matrices with the Standard-Model chiral couplings;
from the degree-thirteen Jarlskog structure onward (a
multi-loop effect), corrections are possible and are at
the Standard-Model level, ≲ 10−16 (App. AO). On the
product background the Dirac operator has no zero
modes, so the anomaly phase is trivial for either a
determinant or a Pfaffian fermion measure.
Corollary AH.67 (Neutron and Electron EDM Predictions). dDFD
∼ (0.9–5.6) × 10−32 e·cm (the CKM
n
long-distance ledger of the “Neutron-EDM ledger and falsifier” corollary in App. AT of the main paper; ∼ 6 orders below the current PSI bound 1.8 × 10−26 e·cm), and
dDFD
≲ 10−38 e·cm (CKM/SM floor, ≳ 8 orders below
e
ACME-III/JILA ∼ 10−30 e·cm).
Proof. With θ̄ = 0 at tree level from Theorem AH.66,
the CP sources are the CKM phase and the loop-induced

h.

Lepton Flavor Violation: Structurally Negligible

Theorem AH.68 (LFV Branching Ratios in DFD). With
MR = MP α3 = 4.744 × 1012 GeV and DFD’s PMNS
matrix structure (order-of-magnitude bounds; the see-saw
rates scale as |Θ|4 ∼ (mν /MR )2 ): BR(µ → eγ) ≲ 8 ×
10−51 , BR(µ → eee) ≲ 2 × 10−55 , BR(τ → µγ) ≲ 3 ×
10−48 , BR(τ → eγ) ≲ 7×10−50 , CR(µ−e, Ti) ≲ 3×10−51 .
Proof. Standard see-saw with heavy Majorana neutrinos at MR gives effective dimension-6 operators
LLFV ⊃ cdipole (mℓ /MR2 ) ℓ̄L σ µν ℓR Fµν + . . ., with the
heavy-neutrino estimate BR(ℓi → ℓj γ) ∼ (3α/32π) |Θ|4
times a PMNS overlap factor, |Θ|2 ∼ mν /MR ∼ 10−23 .
With DFD’s PMNS values and neutrino masses (Theorem AH.28): PMNS overlaps ∼ 10−1 . The various
channels differ by PMNS overlap factors that scale predictably.
Corollary AH.69 (All Planned LFV Experiments Are
Null Predictions). MEG-II (∼ 6 × 10−14 for µ → eγ),
Mu3e (10−16 for µ → eee), Belle II (10−9 for τ channels),
Mu2e/COMET (10−17 for µ-e conversion) all sit ≥ 31
orders above DFD’s predictions. DFD predicts null at
every planned LFV facility. A positive signal at any of
them at level > 10−18 falsifies the see-saw scale MR =
MP α 3 .
Proof. Direct comparison of DFD predictions (Theorem AH.68) with experimental thresholds; gap is ≥ 31
orders in every channel.

23
i.

Galaxy Formation in DFD: the High-z Halo Abundance

On the adopted carrier (Qχ = 1) DFD’s halo abundance at z = 6–15 is Planck ΛCDM’s, lower by 3–19% at
1010 –1011 M⊙ because σ8 and the small-scale power are
lower (Sheth–Tormen, computed). DFD’s halo growth
therefore produces no excess of early galaxies. The luminous galaxies at z ≃ 14 need a star-formation efficiency
of 0.13–0.25 of their halo baryons, about 4% more than
in ΛCDM, and a confirmed population at z ≃ 14 with
n ≥ 10−6 Mpc−3 and M∗ > 4 × 109 M⊙ would exceed the
baryons of every DFD halo. The deep-MOND apparentamplitude branch (Q ≃ 502) gives a 10–30× excess, but
it is not DFD’s growth carrier, and its excess is not a
prediction.
If the epoch law holds on the bound-matter clock, compact galaxies enter the MOND regime early: a 1010 M⊙
galaxy within 1 kpc has ν = 1.6 at z = 6 and 2.2 at z = 14
(computed), which deepens its potential and shortens its
free-fall time by ν −1/2 . Whether this raises the starformation efficiency needs hydrodynamics; each 30% gain
in efficiency raises the counts of bright galaxies at z = 8–
14 by a factor 2–2.5. Kinematics of compact galaxies at
z ≥ 6 decide it: the epoch law predicts Mdyn /Mbar ≥ 1.5
inside a kiloparsec from the boost alone.

j.

Type Ia Supernovae and the Hubble Diagram Plateau

Theorem AH.70 (Retired CMB-anchored Type-Ia
bridge forecast). The former +0.59 ± 0.05 mag forecast
was conditional on the CMB-anchored bridge of Theorem AH.85. That bridge is retired by the exact conformalinvariance and homogeneous-mode results of App. ZS; this
number is not a current DFD prediction. Inhomogeneous
distance effects remain an open numerical program.

k.

Big Bang Nucleosynthesis: Standard Abundances

Theorem AH.71 (DFD-BBN Equivalence to SM-BBN).
DFD predicts the standard primordial abundances:
D/H = 2.55 × 10−5 (vs. obs 2.547 ± 0.025 × 10−5 ), Yp =
0.2479 (obs 0.2453 ± 0.0034, 0.7σ), 7 Li/H ≃ 4.7 × 10−10
(obs 1.6 ± 0.3 × 10−10 ). Neff = 3.044 (obs 2.99 ± 0.17):
the species count is Nν = 3 with no dark radiation (see
proof ), the count itself being the generation count, a discrete input of the main volume (App. F), the residual
+0.044 being the standard SM thermal correction adopted
as input. DFD does not solve the lithium-7 puzzle.
Proof. At BBN (T ≈ 1 MeV), DFD’s optical metric on
FRW background reduces to the standard FRW (āext = 0
by Theorem AE.1). The ratio of every nuclear and weak
rate to the Hubble rate is unchanged from ΛCDM at
T ∼ MeV scales (main volume, App. ZS: gravity enters
as λ2 G and the early rates run faster by λ). Baryon density Ωb h2 is fixed by ηB = 6.1 × 10−10 , whose α4 power

is derived and whose coefficient is open (Thm. AH.33,
§AH 4 c); BBN here uses the observed value, so the numerical BBN inputs are the standard ones. Effective neutrino species: the integer count is the generation count,
Nν = 3, a discrete input of the main volume that coincides with χtop (CP 2 ) = 3, with zero extra relativistic
degrees of freedom. The absence of dark radiation is
triple-locked — Berezin–Toeplitz/Spinc Hilbert saturation
(dim Hmicro = 60 = 3 + 57), Theorem AH.64 (no sterile ν,
dark photons, or ALPs), and Theorem AH.225 (no light
Goldstone mode) — so DFD predicts the absence of dark
radiation as a theorem, not an assumption. The residual thermal correction (+0.044, giving Neff = 3.044 in
the modern Akita–Yamaguchi/Bennett evaluation; legacy
value 3.046) is inherited from standard FRW/BBN thermodynamics — DFD’s optical-metric dynamics are gated
off at T ∼ MeV by Theorem AE.1 — and is not derived
natively. Standard BBN code (PArthENoPE/AlterBBN)
with these inputs reproduces the cited abundance values.
Lithium-7: with ηB at the observed value, DFD inherits
the standard BBN 7 Li tension; stellar-depletion remains
the leading resolution, exactly as in ΛCDM.
Falsifier: a Simons Observatory Neff that departs
from 3.044 by 3σ or more falsifies the 3-generation count
(Neff = 3.044 with Ngen = 3); at the survey’s forecast
precision, σ(Neff ) ≃ 0.05–0.07, that is a departure of
about 0.15 to 0.2; the count is not a tunable parameter:
it is the generation count, a discrete input of the main
volume (App. F), and DFD adds no dark radiation to it.
The α4 power of the baryon density ηB and its B−L bias
are derived, and the printed mechanism gives about (2–
3)×10−9 , three to five times Planck’s value (Thm. AH.33).
l.

NANOGrav PTA: SMBH-Binary Spectrum, Cosmic
Strings Excluded

Theorem AH.72 (NANOGrav 15-yr Compatibility).
DFD predicts a stochastic GW background spectral index
γDFD = 13/3 ≈ 4.33 from supermassive-BH binary inspirals (the only DFD-permitted source). Hellings–Downs
angular correlation ζHD (θ) is the structural GR identity
by cT = c (Theorem AH.20). Cosmic-string γ = 5 is
forbidden (Theorem AH.64, item vii).
Proof. By the Phinney population integral, the stochastic
background from SMBH binary inspirals has characteristic
strain hc (f ) ∝ f −2/3 , giving spectral index γ = 13/3
in the standard PTA convention. DFD’s black holes
are Kerr’s (argued beyond first order in the spin), and
their inspirals are general relativity’s at PTA frequencies
(App. WS). NANOGrav 15-yr observed γ = 3.2 ± 0.6,
statistically consistent with 13/3 = 4.33 at the 1.9σ level.
Cosmic strings would give γ = 5, structurally forbidden
by Theorem AH.64.
Falsifier: SKA1-MID PTA by ∼ 2032 reaches σ(γ) ≈
0.15. A converged γ near 5 at > 3σ falsifies item (vii) of
Theorem AH.64 by demonstrating cosmic strings.

24
m.

Stochastic ψ-PDE: Pathwise Existence and Spectral Gap

Theorem AH.73 (Pathwise Unique Mild SPDE Solution). Let U ≡ (ψ, χ) on R3 × C60 obey the Parisi–Wu
Langevin equation
∂τ U = −Q∇U SE [U ] + η(x, τ ),
with η a Wiener process whose covariance Q is trace class
on H s (R3 ). For s > 5/2, there exists a pathwise unique
mild solution U ∈ C([0, T ]; H s (R3 ) ⊕ C60 ) for any initial
condition U (0) ∈ H s .
The Picard map Φ(U )
=
Sτ U0 −
RProof.
τ
S
Q∇
S
[U
(s)]
ds
+
Z(τ
),
where
S
is the
τ
−s
U
E
τ
0
analytic semigroup of the linearized
operator
and
Z(τ )
Rτ
is the stochastic convolution 0 Sτ −s dη(s). Strong
monotonicity of µ gives the Lipschitz constant C1 (1 + R)4
on bounded balls. Banach contraction on C([0, T ]; H s )
for T < (2C1 (1 + R)4 )−1 gives local pathwise unique
solution; iteration globalizes.
Theorem AH.74 (Invariant Measure and Spectral Gap,
conjectural). On a finite lattice regularization with the
constant mode of ψ fixed (not proved in the continuum),
the Parisi–Wu generator has the invariant Gibbs measure
−1 −SE /ℏ
µβ = ZE
e
dU and, on bounded balls of radius R,
spectral gap γP W ≥ ℏ/µmax with µmax = max(C1 (1 +
R)4 , Λ2top ).
Proof. By Bakry–Émery curvature criterion applied to
the Hessian of SE on H s (R3 ) ⊕ C60. The convexity of
√
√
W (X) = X − 2 X + 2 ln 1 + X (W ′′ (X) > 0 for
X > 0; App. U, main volume) does not by itself give a
positive lower bound, since W ′ (0) = 0; the gap, and with
it total-variation convergence to µβ at exponential rate
γP W , is conjectural. Substituting the lattice parameters
at L = 24 predicts integrated autocorrelation time τint ≈
16, matching the lattice MC specification within constants
of unity.
n.

Modular Tensor Category: the SU (2)58 Chern–Simons
TQFT

Theorem AH.75 (DFD Microsector and the RT TQFT).
The dimension of the DFD microsector equals the shifted
level of the Reshetikhin–Turaev SU (2)58 Chern–Simons
TQFT, a unitary modular tensor category with 59 simple
objects, shifted level 60 = k + h∨ = 58 + 2, and central
charge c = 174/60 = 2.9.
Proof. The Chern–Simons partition function on S 3 at
level k = 58 for SU (2) has 59 irreducible integrable representations labeled j = 0, 1/2, 1, . . . , 29 (i.e. k + 1 = 59 simples). The shifted level is 60 = k +h∨ , where h∨ = 2 is the
dual Coxeter number of SU (2). The dimension of DFD’s
microsector Hilbert space Hmicro = H 0 (CP 2 , O(9)⊕O⊕5 ),
60, equals the shifted level k + h∨ , not the number 59
of simple objects. The main volume quantizes χ’s flux

mode at the level k = 60 itself (App. AV), a different level.
Central charge: c = (3k)/(k + 2) = 3 · 58/60 = 2.9.
Corollary AH.76 (Four Readings of kmax = 60). kmax =
60 admits four readings: (a) Spinc index χ(CP 2 , E); (b)
shifted level k + h∨ of the SU (2)58 MTC; (c) the order
|A5 | of the icosahedral group; (d) |E8 roots|/4.
Proof. Each is a direct invariant computation. These are
readings of the truncation kmax = 60; the main volume’s χ
relic uses the Chern–Simons level k = 60 itself, a different
theory with 61 sectors and K = k + 2 = 62 (App. AV),
and route (v) of Theorem AH.206 names a third, a U (1)60
Chern–Simons invariant.

o.

Spectral Data of DFD

Theorem AH.77 (DFD Spectral Data). DFD carries
the spectral data (ADFD , HDFD , DDFD , JDFD , γDFD ) with:
ADFD = C ∞ (M ) ⊗ M60 (C); HDFD = L2 (M, SM ) ⊗ C60
(chirality-doubled to C120 ); DDFD = DM ⊗ 1 + γ5 ⊗ DF
with DF Spinc Dirac on CP 2 × S 3 at kmax = 60; KOdimension 4 (mod 8).
Proof. Algebra: M60 (C) is the operator algebra of
Hmicro = H 0 (CP 2 , O(9) ⊕ O⊕5 ); tensored with C ∞ (M )
gives the full DFD algebra. Hilbert space: L2 (M, SM )
is the standard spinor Hilbert space; C60 is the microsector
space Hmicro . Dirac: DF has 3 zero modes (corresponding to Ngen = 3) and 57 positive eigenvalues; spectrum
{0, 0, 0, λ4 , . . . , λ60 }. Real structure J: J = JM ⊗ JF
with J 2 = −1 (KO-dim 4). Axioms: order zero fails,
since on the chirality-doubled C120 the commutant of
M60 (C) is M2 (C), which cannot contain JAJ −1 (main
volume, App. AD), and order one presupposes it; orientation γ = c(vol) verified by γ5 identification; regularity
a, [D, a] ∈ ∩n Dom(δ n ); finiteness dim H < ∞ over algebra (true on microsector factor); Poincaré duality, Ker(D)
realizes the K-homology fundamental class. The inner
fluctuations admitted as physical are those of End(E),
which carry the gauge fields (their group is selected, not
forced: Theorem AH.214, Remark AH.215); the 60 eigenvalues of DF are held fixed by the rigidity premise, layer 7
of Theorem AH.179.

p.

Reionization in DFD

On the adopted carrier DFD’s reionization history is
ΛCDM’s at equal astrophysics. In a one-zone model with
the ionizing efficiency set to give τ = 0.054 in Planck
ΛCDM, DFD gives τ = 0.0538 and a midpoint later by
∆z = 0.04 (computed), below the reach of any planned τ
measurement. If the Bose response acts on gas in early
halos with the epoch law a0 (z) on the bound-matter clock,
halos down to ∼ 5 × 105 M⊙ cool atomically and τ rises to
0.078 at the same efficiency, so Planck’s τ = 0.054 ± 0.007
then requires about half the ΛCDM ionizing efficiency, a

25
quantity uncertain by more than a factor ten (computed).
The optical depth is an astrophysical calibration in DFD,
as in ΛCDM; the value τ = 0.061 of the Q ≃ 502 branch
is not a prediction.

q.

Background–Perturbation Decomposition and Scale
Partition

Remark AH.78 (Status note: the “Frame Theorem” label). This subsubsection is the decomposition called the
“Frame Theorem” elsewhere. The two-piece background–
perturbation split is useful, but the cosmological deepMOND carrier (Q ≈ 502.24, F8 = 77.9, σ8 = 0.811)
is an optical/apparent-amplitude construction, not the
physical carrier. The physical growth carrier is χ with
Qχ = 1, conditional on the adopted Rest-Mass Channel,
and the measured-clock forward run gives σ8 = 0.8063
and S8 ≃ 0.822.
Theorem AH.79 (Background-Perturbation Decomposition (two-piece)). On a strictly homogeneous-isotropic
FRW geometry sourced by ρ ≡ ρ̄(t), the DFD scalar field
admits the gauge-fixed split ψ(x, t) = ψ̄(t) + δψ(x, t), with
∇ψ̄(t) ≡ 0 pointwise and aFRW
≡ (c2 /2)|∇ψ̄| = 0 as an
ext
action-level identity. The Euler–Lagrange equation for
ψ̄(t) is identically satisfied for any temporal profile (the
FRW background is a flat direction in the action; the temporal sector is closed separately by the deviation invariant
∆ = (c/a0 )|ψ̇ − ψ̇0 |, App. Q).
Proof. Direct from App. AE Theorem AE.1. The matter
source vanishes by homogeneity (ψ(ρ− ρ̄) = 0 since ρ ≡ ρ̄);
the spatial kinetic vanishes by isotropy (ψ̄ depends only on
t, so ∇ψ̄ ≡ 0); the temporal kinetic vanishes by reference
invariance (∆ = 0 on the background flow that defines
ψ̇0 ). The two-piece split ψ = ψ̄ + δψ is gauge-fixed by the
FRW homogeneity of the matter source, with no further
structural ambiguity.
Assumption AH.80 (Scale-Partition Hypothesis for
δψ Response). The perturbation δψ(x, t) admits a scaledependent response prescription separated by a scale
parameter ksplit (corresponding to a smoothing radius
Rsplit ∼ 100 kpc–few Mpc):
• Cosmological optical-amplitude regime (k ≲
ksplit ): evaluating the operator at x̄ = 0 gives
QDFD ≈ 502. This regime does not govern the
physical σ8 closure; the adopted carrier is χ with
Qχ = 1.
• Galactic-EFE regime (k ≳ ksplit ): the response
operator at a galaxy location xG is evaluated with
operative external acceleration |∇δψ|xG · c2 /2.
The boundary scale ksplit is a modeling choice; the partition is not derived from the action but is the cosmologicalperturbation-theory standard split applied to DFD’s nonlinear response operator. The two regimes use the same

response operator Mij = µ0 δij + µln,0 ĝi ĝj evaluated at
different background gradients, so no new physics is introduced; only the operative ambient |∇ψ| differs.
Remark AH.81 (Status of Assumption AH.80). This is a
named hypothesis in the App. AC sense, not a theorem.
The main volume settles its cosmological side: under the
adopted Rest-Mass Channel the polarization is confined
to bound low-acceleration systems, so unbound linear
perturbations grow with Geff = G at every k, and the
σ8 closure of App. AE is a baryon-only construction,
not DFD’s growth carrier (main volume, App. GR). The
partition therefore does not enter linear growth; it remains
a modeling assumption for the galaxy-scale applications
above.
Theorem AH.82 (Frame Invariance for the Computed Observables). The two local observable classes that
DFD computes from Theorem AH.79+Assumption AH.80
are independent of the bookkeeping choice between the
cosmological-regime and galactic-EFE-regime applications:
1. Galaxy rotation curves are sourced by ∇ϕlocal
(the galaxy’s own perturbation), which is independent of how the cosmological background is gauged.
2. Wide-binary EFE values (App. V.2) are sourced
by the local Galactic gradient |∇ϕGalactic | ≈ 2.0 ×
10−10 m/s2 at the solar neighborhood, independent
of the cosmological-regime application.
Proof. For (1): Theorem AH.79 fixes ∇ψ̄ = 0, so rotation
curves depend only on the galaxy’s interior matter and
the operative external field at xG ; both are defined within
the galactic-EFE side of Assumption AH.80. For (2):
wide-binary gext is dominated by the Galactic potential
gradient at D ∼ 8 kpc, an order of magnitude above
the cosmic peculiar gradient at the solar neighborhood;
cosmological-regime choice does not affect it.
Remark AH.83 (Limit of Frame Invariance). Theorem AH.82 establishes the two local statements above. It
does not extend to arbitrary f σ8 (k). The current physical
growth calculation is instead conditional on the adopted
Rest-Mass Channel; a full survey-forward implementation
remains an open numerical obligation.
r.

∆ψ Growth Bridge

Theorem AH.84 (Retired CMB-screen growth equation).
 2
d∆ψLOS
3
H
= Ωb
R8 Q(z) Db (z)
d ln(1 + z)
2
c
was the equation used by the archived optical branch. It is
not a current physical-growth or acoustic-screen equation.
Proof. Linearize the elliptic operator ∇·[µ(|∇ψ|/a⋆ )∇ψ]
about the FRW background (ā = 0, deep-MOND limit
µ → x). On the line of sight integrated against

26
the spherical-top-hat
R8 window, the contribution is
Rχ
∆ψLOS (z) = 0 (z)dχ′ · 3H02 /(2c2 ) · Ωb R8 Q(z ′ )Db (z ′ ).
Differentiating in ln(1 + z) and using dχ/d ln(1 + z) =
c/H(z) gives the displayed evolution equation. Key
inputs: Q(z) is the deep-MOND amplification (Theorem AH.7), Db (z) is the standard linear baryon
growth.
Theorem AH.85 (Retired CMB-anchored closed-form
bridge). The former piecewise expression with a1 = 0.190,
a2 = 0.0049 and a ∆ψ(1100) = 0.300 CMB endpoint is
retired. Exact conformal invariance leaves the homogeneous endpoint no screen beyond the expansion, while the
computed inhomogeneous mean is orders of magnitude too
small (App. ZS). The local screen program is preserved
but must be rederived without a CMB anchor.

s.

Action-Principle GHY-Analog Boundary Term

Theorem AH.86 (Boundary-Term Closure). The full
DFD action is S = SDFD + Sbdy , with SDFD the complete
action of the main volume.
Proof. Standard Hamilton’s principle: vary ψ in S. Without
term,
R a boundary
R δS on a bounded region Ω produces
4
[EL
]δψ
d
x
+
flux(δψ, ∇δψ) dS. The boundary inψ
Ω
∂Ω
tegral doesR not vanish for arbitrary δψ at ∂Ω. Adding
Sbdy = − ∂Ω K eψ nµ ∇µ ψ dS (Gibbons–Hawking–York
analog with K = trace of extrinsic curvature
R on the slice)
cancels the boundary flux, giving δS = Ω [ELψ ]δψ d4 x.
The variational principle is fully consistent without imposing δψ|∂Ω = 0.
7.

CKM, Higgs Self-Coupling, Halo Profiles, and
Anomaly Cancellation

Every theorem in this subsection is followed by an
explicit proof. Independent recomputations of TPI and
CKM are recorded as status notes beside the results they
grade.

a. CKM ρ̄ Cayley-Graph Half-Integer Correction
(Conjectural Selection-Rule Extension, Not a Derived
Theorem)

a. Canonical treatment: Appendix AO. DFD’s CPsector result is the consolidated treatment of Appendix AO. There the magnitude skeleton λ = 31α,
A = 108α, η̄ = 49α is built, with selected integers,
from line-bundle cohomology on CP 2 together with
D = dimR (CP 2 × S 3 ) = 7, and the CP-even apex is
fixed at ρ̄ = 43
2 α = 21.5α by the Euler-projection postulate (ρ̄/α = D2 /2 − χ(CP 2 ) = 49/2 − 3). The Euler
projection replaces the post-hoc “19 + 5/2” half-integer
Cayley correction recorded below: it lands on the same

Path A value 21.5α but from a structural selection principle rather than from a correction chosen to preserve the
integer 19. The weak CP phase is treated at theorem
grade in Appendix AO: with real (Hermitian) Yukawa kernels the literal flavor construction localizes the generations
and the real Higgs profile on the Lagrangian real locus
RP 2 ⊂ CP 2 , the CKM matrix is real orthogonal, and the
Jarlskog invariant vanishes (J = 0, Theorem AO.7); a single conjugation-odd, determinant-orthogonal Berry offset
(Postulate Y.10) then supplies the observed phase while
keeping θ̄ = 0 at tree level, giving the unitarity-triangle
angle γ = arctan(98/43) = 66.3093◦ at theorem grade
inside the strengthened DFD–SD branch. The conjecture
γ = 4π/11 is not DFD’s prediction. The integer-pattern
analysis below serves the downstream flavor-observable
inputs (B-mixing, ε′ /ε, rare kaon decays), which take the
apex as input.
Remark AH.87 (Status of the CKM ρ̄ “closure”). The
construction below is a Candidate Selection-Rule Extension, not a Theorem, because of its post-hoc character.
The +5/2 correction was motivated by drift in the PDG
ρ̄ central value (0.141 in 2018–2020 → 0.159 in 2024),
not derived ab initio from earlier DFD axioms. The three
“independent” arguments (Cayley graph, Verlinde label,
Connes Dirac) share the same underlying A5 +bundle
combinatorics on {O(1), O(2), O(3)} — they are three
descriptions of the same object, not three independent
confirmations. The integer 22 (giving 0.91% vs PDG
2024) is closer than 21.5 (1.39%), but 21.5 = 19 + 5/2
was selected to preserve the original integer 19 rather
than rebuild the bundle assignment. We record the construction here as a candidate selection-rule extension,
with explicit acknowledgment of its post-hoc status, pending structural work that could either (a) derive +5/2
from a forcing condition independent of the data drift,
or (b) accept the alternative ρ̄ → 22α with bundle triple
{h0 (O), h0 (O(2)), h0 (O(4))} = {1, 6, 15} as the corrected
integer assignment. Until then this is a postdiction, not
a prediction.
Proposition AH.88 (Candidate ρ̄ Selection-Rule Extension — conjectural). Conditional on the original integer
pattern (31, 108, 19, 49) being preserved, a half-integer
Cayley correction δρ̄ = +5/2 exists that gives
ρ̄cand
DFD = (19 + 5/2) α = 21.5 α = 0.15689,
matching PDG 2024 ρ̄ = 0.1591 ± 0.0094 at 1.4%. Within
the same conditional framework, λ, A, η̄ receive δ = 0
corrections, preserving their integer-pattern values. This
is a candidate extension, not a derived theorem;
see the status remark above.
Remark AH.89 (Alternative: integer-22 reassignment).
If the bundle triple is allowed to be reassigned, ρ̄ →
22α = 0.1605 matches PDG 2024 at 0.91% and corresponds to bundle triple {1, 6, 15} (sums to 22) instead
of the original {3, 6, 10} (sums to 19). This alternative
requires retiring the original 19 identification. The choice
is between (a) the 19 + 5/2 epicycle preserving integer

27
19 and (b) the integer-22 reassignment, and neither is
derived from forcing axioms; the Euler-projection postulate of Appendix AO reaches the value of (a), 21.5α, by
a different route.
Plausibility argument (not a theorem-grade proof ).
The CKM mixing originates from off-diagonal
connectors in the Yukawa Berry-bundle on CP 2
(Theorem AH.27). The Wolfenstein integer pattern
(nλ , nA , nρ̄ , nη̄ ) = (31, 108, 19, 49) encodes hop distances
on the Cayley graph (A5 , S). Within the assignmentsas-given, three conditions can be imposed that single
out ρ̄ for a half-integer correction: (i) bundle-orbit triple
containing O(1); (ii) first-generation projector image;
(iii) Z3 -fixed orbit. Conditional on these conditions and
on the original bundle assignment, δρ̄ = +5/2 follows.
The three “independent” arguments (Cayley graph hop;
Verlinde half-integer label at the apex; Connes Dirac
off-diagonal element) all reduce to the same combinatorial
datum (A5 , {O(1), O(2), O(3)}) and are not independently derivable; threefold convergence here reflects
shared assumptions, not independent confirmations. An
independent check graded this confirmed with caveats:
numerically excellent at 1.4%, η-asymmetry test passes,
but +5/2 is a best-fit-among-allowed-half-integers, not a
forced value. Theorem-grade promotion requires deriving
the conditions (i)–(iii) themselves from a forcing axiom
independent of ρ̄ data.
Corollary AH.90 (The candidate apex pattern). The
CKM integer pattern is (λ, A, ρ̄, η̄) = (31, 108, 21.5, 49) α.
Numerical predictions: λ = 0.226, A = 0.788, ρ̄ =
0.15689, η̄ = 0.358. Mean relative error vs PDG 2024:
∼ 2%. Worst channel: A at 4.6% (independent of the ρ̄
correction). Unitarity-triangle apex angle γDFD = 66.3◦
DFD
vs PDG 65.4◦ ±1.5◦ (1σ). Jarlskog JCP
= 2.9731×10−5
−5
(main volume, App. AO) vs PDG (3.12+0.13
−0.12 ) × 10
(1.2σ).
Proof. Substituting ρ̄ = 21.5α into the standard Wolfenstein parameterization with α = 7.297 × 10−3 yields the
displayed values. Computing the unitarity-triangle apex:
tan γ = η̄/ρ̄ = 0.358/0.15689 = 2.282, hence γ = 66.3◦ .
Jarlskog: J = η̄ A2 λ6 = 0.358 · 0.621 · 1.34 × 10−4 =
2.98 × 10−5 at leading order.
Falsifier: LHCb Run 3 + Belle II by 2030 reach σ(ρ̄) ≈
0.005, σ(γ) ≈ 0.3◦ . DFD falsified at 3σ if measured
ρ̄ ∈
/ [0.143, 0.171] or γ ∈
/ [65.4◦ , 67.2◦ ].
Remark AH.91 (Per-channel CKM agreement). Against
PDG 2024 (ρ̄obs = 0.1591 ± 0.0094), the bare integer pattern (31, 108, 19, 49)α gives mean relative error ∼ 5%,
with per-channel breakdown: λ at 0.5% (1.8σ), η̄ at
1.5% (0.7σ), A at 4.6% (2.5σ, sensitive to the open inclusive/exclusive |Vcb | tension), ρ̄ at 12.9% (2.2σ) (the raw
ρ̄ = 19α residual; the Euler-projected apex ρ̄ = 43
2 α of
App. AO in the main paper lands at ∼ 0.2σ against PDG
2024). The abstract of the main volume reports
this per-channel breakdown. Proposition AH.88 (the

+5/2 Cayley correction) is recorded as a conjectural candidate selection-rule extension, not a derived theorem; see
the status remark immediately preceding it. The integer
pattern itself is suggestive of cohomological origin; the
selection rule remains open.

b.

Pion Decay Constant from GMOR

Theorem AH.92 (fπ from Per-Flavor GMOR with DFD
Inputs). In the Fπ ≃ 92 MeV normalization, per-flavor
GMOR with the series IV–VI condensate estimate gives
s
(mu + md ) Σ
Fπ
DFD
×
fπ
=
m2π
F0 NLO
= 87.05 × 1.064 = 92.6 ± 3.7 MeV,
¯ is the per-flavor chiral conwhere Σ ≡ |⟨ūu⟩| = |⟨dd⟩|
densate, Σ1/3 = 272 ± 26 MeV (Theorem AH.132, with
the error of ΛQCD,3 ), and the quoted ±3.7 MeV (4.0%)
propagates the FLAG condensate band (272 ± 5 MeV);
the fully DFD-internal ±26 MeV band gives ±13 MeV
(14%). The result sits +0.4% from PDG 92.28 MeV —
0.03σ on the internal budget, 0.09σ on the FLAG band.
Precision claim: 3–5% (FLAG band) and 14% (fully internal), consistent with Corollary AH.133. Its inputs are
mu + md , which rests on data-selected prefactors (main
volume, App. FM), the selected strong scale, and the
FLAG Nf = 2+1 chiral-log factor Fπ /F0 = 1.064(7).
Proof. Per-flavor GMOR in the 92-normalization reads,
at leading order, Fπ2 m2π = (m√
u + md ) Σ. (In the fπ ≃
130 MeV convention, fπ = 2 Fπ , the same relation
carries the factor 2: fπ2 m2π = 2(mu + md )Σ — the factor 2
belongs to the 130-convention only.) Inputs, all from the
DFD ledger: mu +md = 6.86±0.3 MeV (Theorem AH.27,
from Au , Ad prefactors and αnu , αnd exponents); Σ1/3 =
272 ± 26 MeV, the series IV–VI condensate estimate
(Theorem AH.132, from ΛQCD,3 = 332 ± 20 MeV; FLAG
2024 lattice: 272 ± 5 MeV), which replaces the NDA value
−(243±12 MeV)3 built on ΛQCD,3 = 210±14 MeV. With
mπ0 = 134.977 MeV:
p
(6.86 MeV) (272 MeV)3
Fπ (LO) =
= 87.05 MeV.
134.977 MeV
The NLO chiral-log correction is the standard ratio
Fπ /F0 = 1.064(7) (FLAG Nf = 2+1 average), giving
fπ = 87.05 × 1.064 = 92.6 MeV. Error budget: ±3.7 MeV
(4.0%) propagating the FLAG condensate band, ±13 MeV
(14%) propagating the DFD-internal ±26 MeV band;
σ(mu + md ) = 0.3 MeV and the NLO ±0.007 are subdominant. Against PDG 92.28 MeV the deviation is +0.4%
(0.03σ internal, 0.09σ FLAG band).
√
The naive NDA estimate fπ ∼ 4πΛQCD / 6 that produces ∼ 326 MeV (the so-called “fπ puzzle”) confuses
dimensional analysis with the actual GMOR relation; once
the chiral condensate is included with its proper scale, the
fπ value drops to ∼ 92 MeV with no extra suppression
mechanism required.

28
Remark AH.93 (Status note: the NDA-condensate fπ
arithmetic does not close). The evaluation fπ2 = 2 · 6.86 ·
(243)3 /(135)2 = (92.0 MeV)2 does not hold: direct evaluation of that expression gives (103.9 MeV)2 — and
(73.5 MeV)2 without the factor 2 — so 92.0 does not
follow from those inputs under either GMOR normalization. The present theorem therefore disentangles the two
normalization conventions (the factor 2 belongs to the 130convention), uses the series IV–VI estimate −(272 ± 26)3
(Theorem AH.132) in place of the −(243)3 NDA condensate, and applies the NLO chiral-log factor explicitly.
The precision claim is the 3–5% (FLAG band) / 14%
(fully internal) of Corollary AH.133, not “0.5% relative
precision”.
Falsifier: Roy-Steiner combination 2a00 −5a20 = 0.662±
0.012 m−1
π (DFD prediction) vs. E865+DIRAC 0.6437 ±
0.0144 m−1
π . NA62 final Kℓ4 + LHCb amplitude analysis
reaches ±0.005 m−1
π by 2030 — DFD falsified if measured
outside [0.62, 0.71].

c.

Higgs Self-Coupling κλ

Theorem AH.94 (Higgs self-coupling). With the Higgs
quartic fixed at a matching scale (the conjectured 1/8,
main volume App. Z), DFD’s trilinear self-coupling is the
Standard-Model one at the resulting Higgs mass: κλ = 1
at tree level.
SM
Proof. κλ ≡ λobs
HHH /λHHH . With the quartic fixed at
a matching scale, DFD’s Higgs sector is the Standard
Model’s at the resulting mass, so the two coincide. A ratio
of the conjectured 1/8 to the physical quartic measures the
tree-to-pole mass shift and is not a coupling modifier, and
a boundary λH (MP ) = 1/8 would give λH (2mH ) = 0.174
under two-loop running (computed), far from the physical
quartic.

HL-LHC σ(κλ ) ≈ ±0.5 by 2030 and FCC-hh ±0.05
by ∼ 2050 measure the Standard-Model value that DFD
shares; they do not discriminate.

d.

Strict Lepton-Flavor Universality and B-Meson
Predictions

(G2F /6π 2 ) ηB m2W S0 (xt ) |Vtq |2 BBq fB2 q mBq . The DFD
apex gives, at LO Wolfenstein, (1 − ρ̄)2 + η̄ 2 = 0.839
(Theorem AH.88), i.e. |Vtd /Vts |2 ≈ 0.0429; the exactmatrix evaluation of the same locked apex (App. AO,
|Vtd | = 0.008358, |Vtd /Vts | = 0.2109) gives |Vtd /Vts |2 =
0.0445, which replaces the LO shortcut. Combined
with the SU(3)-breaking ratio ξ 2 = BBs fB2 s /BBd fB2 d =
1.4544 (FLAG ξ = 1.206(17)) and the B-meson mass
ratio mBs /mBd = 1.01653, this gives ∆Ms /∆Md =
ξ 2 (mBs /mBd )/|Vtd /Vts |2 = 1.4544 · 1.01653/0.044487 =
33.23 (−5.1% vs PDG; ≈ 1.9σ with the ξ 2 error dominant
— see the tension box below). The γ angle follows from
the unitarity triangle (Corollary AH.90). fBs from the
same chiral-condensate structure as fπ (Theorem AH.92),
with strange-quark mass correction.
Remark AH.96 (Tension Box: εK and ∆Md /∆Ms under the Locked Apex). The locked tower (λ, A, ρ̄, η̄) =
(31, 108, 43
2 , 49)α has no free Wolfenstein parameters; its
two mixing tensions decompose as follows. (i) No native
amplitude: DFD adds no new ∆F =2 term (χ, ψ couple
flavor-diagonally; gravity box ∼ (mb /MP )4 ≈ 10−74 ; αconnector dressing ≤ 0.7%), so DFD mixing is the SM
box on the locked CKM. (ii) εK : |Vcb | = Aλ2 = 0.04033
sits on the exclusive side of the |Vcb | puzzle (+2.1% vs
excl. 39.5, −4.4% vs incl. 42.2×10−3 ). With εK ∝ |Vcb |3.4
(B̂K =0.7625, κε =0.94, NLO ηi ; Im λt = 1.3493 × 10−4
from the exact-matrix apex), DFD gives 1.900 × 10−3 vs
2.228 × 10−3 (−14.7%, −1.62σ; −10.3% under the Buras–
Guadagnoli–Silvestrini normalization). Control: the SM
with a free PDG apex at the same |Vcb | gives 1.90 × 10−3
(−1.8σ) — DFD’s εK coincides with its own SM control;
the pull is inherited |Vcb | normalization, not the apex.
Falsifier: Belle II |Vcb | at 0.5%: inclusive confirmation
leaves A = 108α wrong by 4.4% with nothing to move
(A ≃ 113α forbidden by nA = 3 h0 (O(7)) = 108); DFD
survives only on |Vcb | ≲ 41 × 10−3 . (iii) ∆Md /∆Ms : the
exact-matrix |Vtd /Vts | = 0.2109 (App. AO; not the LO
0.2072) gives ∆Ms /∆Md = 33.23 at FLAG ξ = 1.206(17)
vs 35.07 (−5.1%); with ξ 2 (±2.8%) dominating the budget
(experiment ±0.4%, DFD-locked 0%), the pull is ≈ 1.9σ,
not +4.25σ (which required σξ 2–4× below FLAG). |Vcb |
cancels here; DFD is +2.7% above the ∆M -extracted
0.2053(29) but ∼ +1% above the SM angles-driven fit
0.2087. Decider: FLAG ξ at 0.5%; reconciliation needs
ξ ≃ 1.24 (+1.9σ above FLAG). Status: inherited/latticelimited tensions with no fitted rescue available by construction; decided externally by the two falsifiers above.

Theorem AH.95 (B-Meson Mass-Splitting Ratio).
∆Ms /∆Md = 33.23 (DFD, exact-matrix |Vtd /Vts |, FLAG
ξ = 1.206(17)) vs. PDG 35.07 ± 0.4 (−5.1%; ≈ 1.9σ once
the FLAG ξ 2 error of ±2.8% — the dominant budget item
— is propagated; see Rem. AH.96. The LO Wolfenstein
value is |Vtd /Vts |2 = 0.0429; the exact-matrix value from
the same locked apex is 0.0445.) γ apex angle = 66.3◦ vs.
PDG 65.4◦ ± 1.5◦ (+1.4%, within 1σ). fBs = 224 MeV
(DFD) vs. lattice 230.3 ± 1.3 MeV (−2.7%; 4.8σ below on
the lattice error alone, with no DFD error computed).

Corollary AH.97 (RK and P5′ “Anomalies” Are Not
New Physics in DFD). DFD predicts strict lepton-flavor
universality (LFU): RK ≡ BR(B → Kµµ)/BR(B →
Kee) = 1 up to the Standard Model’s own corrections. The P5′ angular-analysis tension is interpreted as a
hadronic charm-loop matrix-element systematic, with no
Wilson-coefficient shifts from new physics.

Proof. The ∆Mq ratio is the standard QCD boxdiagram with internal top-quark loops, ∆Mq =

Proof. The (3,2,1) gauge bundle in DFD’s microsector decomposition couples leptons to gauge bosons via the same

29
hypercharge Y and SU(2)L generator structure regardless
of generation index. Hence the lepton-flavor-violating
µµ
ee
Wilson coefficients C9µµ − C9ee and C10
− C10
vanish identically by gauge symmetry: RK = 1 exactly at tree level,
with corrections O(m2µ /m2b ). The LHCb 2023 update of
RK = 0.949 ± 0.046 (consistent with unity) supports this
prediction. P5′ depends on long-distance hadronic matrix
elements involving charm-quark loops; its ∼ 3σ persistent
low-q 2 deviation is structurally a QCD systematic, not a
new-physics signal.
Falsifier: LHCb Run 4 RK to σ < 0.005 — any persistent deviation from the Standard-Model value falsifies
DFD’s gauge-bundle LFU.

e.

The Baryons’ Phantom Halo Profile: Flat Cores

Theorem AH.98 (DFD Effective Halo Profile). For a
galaxy with baryonic mass Mb and characteristic radius
rb , the DFD-effective “dark-matter-equivalent” density
profile is


(
1
ρ
(r)
·
−
1
→ 0 r ≪ rM
b
µ(x)
√
ρDFD
eff (r) =
GMb a0
r ≫ rM
4πGr 2
p
with topological transition radius rM =
GMb /a0 ≈
3.41 kpc (Mb /1010 M⊙ )1/2 on bound-matter clocks (main
volume, App. ZS). The profile transitions from “vanishing
halo” inner regime to isothermal-sphere (ρ ∝ r−2 ) outer
regime at r ∼ rM .
Proof. By Poisson’s equation in DFD’s optical-metric
framework: ∇2 ΦDFD = 4πG ρeff , where ΦDFD is the effective Newtonian potential and ρeff = (1/4πG)∇2 ΦDFD −ρb
subtracts the baryon component. The DFD potential satisfies ∇·[µ(|∇Φ|/a0 )∇Φ] = 4πGρb (Theorem AH.21 with
µ = x/(1 + x)).
Inner regime r ≪ rM (deep-Newtonian, |∇Φ| ≫ a0 ):
µ → 1, ∇2 ΦDFD → 4πGρb , hence ρeff → 0. The phantom
dark-matter halo vanishes inside rM .
Outer regime r ≫ rM (deep-MOND, |∇Φ| ≪ a0 ):
µ → x = |∇Φ|/a0 , ∇·[|∇Φ|2 n̂] =√4πGρb a0 . For spherical
M
r: |∇Φ| = GMb a0 /r, hence ρeff =
√ b exterior at large
GMb a0 /(4πGr2 ), i.e. isothermal-sphere ρ ∝ r−2 .
The transition radius isp set by |∇Φ| = a0 ⇒
2
GMb /rM
= a0 ⇒ rM =
GMb /a0 . Plugging a0 =
1.1966 × 10−10 m/s2 and Mb = 1010 M⊙ gives rM ≈
3.41 kpc.
The baryons’ phantom has ρeff → 0 in the inner regime.
This is the phantom of the baryons alone: the derived
χ (main volume, App. AV) is real cold dark matter and
forms halos too, and whether they show in a rotation
curve depends on how χ composes with the µ-law, which
is the adopted floor law (main volume, App. AV).
Corollary AH.99 (Core radii of the phantom). The
phantom’s core radius follows rcore ∝ Mb0.5 in DFD
(from rM scaling), matching the empirical dwarf relation

Mb0.4±0.1 within errors; a rotation curve shows this core
where χ inside the galaxy stays thinner than the phantom.
χ is cold dark matter and forms subhalos under any composition with the µ-law, so DFD does not remove subhalos,
and the missing-satellites and too-big-to-fail questions are
not settled by it.
p
Proof. By Theorem AH.98, rM = GMb /a0 scales as
1/2
Mb . Observed dwarf cores follow Mb0.4 –Mb0.5 with ∼ 0.2dex scatter (de Blok+ 2008, Oh+ 2015, Read+ 2017).
DFD prediction (slope 0.5) is consistent. Subhalos: χ
clusters as cold dark matter on linear scales (Qχ = 1) and
every printed small-scale modifier acts far below galaxy
masses, so χ forms subhalos. Under the printed additive
composition they form as in ΛCDM (computed); how
their nonlinear growth changes under the adopted floor
composition with the µ-law is not computed (main volume,
App. AV).
Stream gaps: χ subhalos perturb cold stellar streams
(Pal 5, GD-1, future Roman/DESI streams) as collisionless
dark-matter subhalos do in ΛCDM, so the gap count
does not separate DFD from ΛCDM unless the floor
composition of χ with the µ-law changes how subhalos
form.
f.

Sunyaev-Zel’dovich Effect and Cluster Counts

Theorem AH.100 (DFD Cluster Y -M Scaling). The
DFD cluster Y -M scaling at z < 0.5 matches ΛCDM at
slope αY = 1.65 ± 0.05 (statistically indistinguishable from
5/3). On the adopted carrier (Qχ = 1) the abundance of
massive halos at z > 1.5 is ΛCDM’s, slightly lower, so
the cluster counts carry no DFD excess.
Proof. Cluster virial equilibrium: Mvir ∝ T σ 2 /Geff
with Geff = G in deep-Newtonian inner regime (TheoremRAH.98). Hence M -T scales as T 3/2 standard. SZ
Y ≡ (kB Te /me c2 )ne σT dl ∝ M T ∝ M 5/3 , slope 5/3 recovered. The high-z abundance follows from linear growth
with Qχ = 1 (main volume, App. GR); the deep-MOND
amplification Q ≈ 502 is not DFD’s growth carrier.
Falsifier: a cluster-count excess over ΛCDM at z > 1.5
at more than 5σ in the SZ surveys of SPT-3G and the
Simons Observatory contradicts the adopted carrier.
g.

Cosmic Shear S8

Theorem AH.101 (DFD S8 prediction on the early–
clock background). With As = 32πα5 , the current inearly
puts
= 67.52, ωb = 0.0222730, ωχ = 0.1192806,
P H0
mν = 61.46 meV, Neff = 3.044, and the normal hierarchy, a forward high-accuracy CAMB evaluation gives
R8 = 8h−1 = 11.848 Mpc, σ8 ≃ 0.8063, and Ωm ≃ 0.312,
hence
p
S8DFD = σ8 Ωm /0.3 ≃ 0.822.

30
This is consistent with KiDS-Legacy (0.815+0.016
−0.021 ) and
with its joint external-probe result (0.814+0.011
−0.012 ). The
linear-clustering statement remains conditional on the
adopted Rest-Mass Channel postulate that supplies Qχ =
1; no separate low-S8 suppression is claimed.
Proof. The smoothing radius changes with the clock
and therefore cannot be frozen: h = 0.6752 gives
R8 = 11.848 Mpc. Running the printed physical densities, normal neutrino spectrum, tilt, and amplitude
through the same transfer and top-hat window gives
σ8 ≃ 0.8063. Including the massive-neutrino contribution
gives Ωm ≃ 0.312, so direct substitution yields S8 ≃ 0.822.
The 0.784 value is the output of a one-clock background
with H0 = 72.09 and Ωm ≃ 0.275 (Apps. GR/CL/PK);
under the two-clock postulate (main volume, App. ZS)
72.09 is the bound-matter clock. The CMB-lensing crosscheck is Aϕϕ = 0.9887 (−0.80σ).
Falsifier: Euclid + LSST joint cosmic-shear precision
near σ(S8 ) ≈ 0.005 directly tests the current S8 ≃ 0.822
forward prediction, with the printed input and CAMBconvention uncertainty propagated rather than a frozen
interval.
h.

Galaxy-Galaxy Lensing and Time-Delay Cosmography

Theorem AH.102 (Galaxy–Galaxy Lensing ∆Σ from
Baryons Alone, where χ Stays below the Demand). Exterior to an isolated baryonic mass Mb , in the deep regime
g ≪ a0 and where the lens’s χ stays below the µ-law’s
demand, DFD predicts the excess surface density
p
Mb a0 /G
∆ΣDFD (R) =
,
4R
equivalentlyp a lensing radial-acceleration relation
glens (R) = gbar (R) a0 with the same a0 that governs
rotation curves: lensing and dynamics trace one optical
potential, with zero additional parameters.
Proof.
The deep-regime exterior field is |∇Φ| =
√
GMb a0 /r (Theorem AH.98), equivalent
to the ef√
fective source density ρeff (r) =
GMRb a0 /(4πGr2 ).
∞
Line-of-sight projection gives Σ(R) = −∞ ρeff dl =
√
√
√
R
Mb a0 /G
GMb a0 ∞
dl
π
b a0
= GM
·R
=
. For any
4πG
4πG
4R
−∞ R2 +l2
RR
2
′
′
′
Σ ∝ 1/R profile, Σ̄(< R) = R2 0 Σ(R ) R dR = 2 Σ(R),
p
hence ∆Σ ≡ Σ̄(< R) − Σ(R) = Σ(R) = Mb a0 /G /(4R).
Because every lensing observable is linear in the amplitude ofpthe 1/r2 effective density, a normalization
∆Σ = x Mb a0 /G/R maps under exact deprojection
√
to glens = 4x gbar a0 ; x = 1/4 is precisely the statement
that weak lensing obeys the dynamical RAR.
Observational status. The published deep-regime
tests are KiDS-1000 galaxy–galaxy lensing around isolated
galaxies (Brouwer et al. 2021, A&A 650, A113; Mistele
et al. 2024, JCAP 04, 020); neither DES-Y3 nor HSCY3 has published a comparable known-Mb isolated-lens

measurement, so no result is attributed to those surveys. Against the spectroscopic GAMA isolated-lens sample, the zero-parameter line passes with full-covariance
χ2red = 0.8 (0.4σ) (Brouwer et al. 2021). Against the
larger photometric KiDS-bright sample the raw fit is
poor (χ2red = 4.6, data ∼ 0.1 dex high), improving to
χ2red = 4.0 within the isolation-criterion limit and to 1.5
under the +0.2 dex stellar-mass systematic (−0.2 dex
gives 14: the data sit high, so only baryon-mass increases
— heavier SPS masses or circumgalactic gas — can reconcile, and both are independently motivated: Mistele
et al. find KiDS early-type masses low by a factor 1.4–
1.7); Brouwer et al. conclude the prediction “describes
our measurements within the statistical and systematic
uncertainties.” Mistele et al. (2024), using consistent
SPS masses and an exact profile-free deprojection, extend the lensing RAR to gbar ≈ 10−14 m s−2 (R ≈ 1 Mpc;
tentatively to 10−14.9 , where their quoted extrapolation
systematic reaches 0.67 dex): against their measured relation the 1/(4R) line gives χ2 /N = 1.06 over the twelve
systematics-clean points (diagonal statistical errors; no
bin-to-bin covariance is published for this table), with
a mild global amplitude residual of +0.04 ± 0.02 dex
(stat) relative to the pure deep-limit line, reducing to
+0.03 ± 0.02 dex against the full interpolation-function
curve (the shallowest lensing bins carry ν-function corrections of up to 0.04 dex); lenses are bound systems, which
read a0 = 1.1966 × 10−10 under the two-clock postulate of
the main volume, App. ZS, the value the residual above
is taken at; the observers’ distance convention (H0 = 73
against 72.09) accounts for log10 (73/72.09) = +0.005 dex,
and about +0.03 dex (+0.02 dex against the full curve) is
left unexplained — inside the ±0.1 dex coherent stellarmass systematic (0.2–0.3σ), and consistent with their
adopted kinematic a0 = 1.24 × 10−10 m s−2 . Normalization discrimination is asymmetric. The value x = 1/4
is the best-fitting of the three values considered and the
unique zero-parameter choice — it is the dynamical RAR.
The value x = 1/π is disfavoured at the statistical level
only (∆χ2 ≃ +10, diagonal errors); once the coherent
stellar-mass systematic is marginalized the two are indistinguishable in current lensing data (∆χ2 ≲ 0.5), and
the case against 1/π rests on cross-consistency: it would
require a lensing-only a0 ≃ 0.9 × 10−10 m s−2 , 26% below
the kinematic value, and a +0.05 dex lensing–kinematic
BTFR velocity offset. The value x = 1/2 is excluded
by three independent channels: per-point residuals of
+0.30 dex (≈ 4σ each, χ2 /N ≈ 19–22; a 2.6σ deficit
survives even after conceding the entire ±0.2 dex stellarmass budget), a required a0 ≃ 3.6 × 10−11 irreconcilable
with rotation curves, and a predicted +0.15 dex offset of
lensing-inferred flat velocities from the kinematic baryonic
Tully–Fisher relation that is not observed: the same lensing data yield “rotation curves” that stay flat to ∼ 1 Mpc
and lie on the kinematic BTFR (Mistele et al. 2024b).
Theorem AH.102 and the fits quoted above treat each
lens as an isolated deep-MOND source with no outer edge.
In DFD the Rest-Mass Channel confines the phantom

31
of each galaxy to its bound region, and the net induced
mass of that region is zero (main volume, App. GR), so
the 1/R profile holds out to about the halo radius r200m
(0.25–0.4 Mpc for Mb = 1010 –1011 M⊙ ; up to 0.6 Mpc if
the bound region reaches 1.5 r200m ). Against the stacked
measurement of Mistele et al., with χ halos lighter than
abundance matching gives, DFD’s profile fits the 7–11 bins
inside that radius and, with the χ halo and the two-halo
term included, falls below the 4–8 bins beyond it (computed). The flat lensing profile to 1 Mpc is therefore not a
DFD prediction: the outer bins are in tension with DFD,
by an amount that depends on how far the bound region
reaches. With χ at abundance-matching halo masses, χ
fills the µ-law’s demand inside the bound radius of discs
above about 1010.9 M⊙ of baryons and ellipticals above
about 1011.2 M⊙ ; there DFD’s inner lensing is ΛCDM’s,
0.1–0.3 dex above the line of Theorem AH.102, while the
stacks’ inner bins lie near that line, so DFD with χ at
those masses overshoots them as ΛCDM’s halo model
does (computed).
Falsifier: the amplitude is locked — no rescaling
freedom. Inside the bound radius of the lenses (about
r200m ), a lensing-RAR measurement with spectroscopic
masses and percent-level stellar-mass control (Euclid +
4MOST/DESI spectroscopy, ∼2027–2030) that finds the
√
deep-regime amplitude offset from gbar a0 (at a0 =
1.1966 × 10−10 , the bound-matter clock of the main volume’s App. ZS) exceeding ∼ 0.1 dex, in lenses whose χ
stays below the µ-law’s demand, falsifies the prediction
(in heavier lenses the prediction is the χ-filled one above);
equally, any confirmed departure of ∆Σ from
p the 1/R
shape well outside the MOND radius rM = GMb /a0
and inside both the bound radius and the isolation limit
(for R ≳ 10 rM the residual baryonic point-mass term
Mb /πR2 contributes < 13%; inside a few rM the 1/R
shape is not predicted).
Theorem AH.103 (Time-Delay Cosmography H0 :
no probe selectivity). DFD’s topological invariant
GℏH02 /c5 = C α57 (Theorem AH.22; coefficient C =
0.87784 ± 0.00266 measured — the clock-credit remark
of App. AP, main paper) sets the early-clock H0early =
67.52 ± 0.09 km/s/Mpc, concordant with the Planck CMB
inference 67.4 ± 0.5. Because DFD’s optical and matter metrics are conformally related, every null-path distance probe rides the same conformal class: a ψ-screen
cannot be probe-selective, and time-delay H0 must agree

with the acoustic channels when both are read in the
same units. The two-clock postulate of the main volume (App. ZS), a ratio between frames and not a screen,
reads the time delays on laboratory clocks and returns the
bound-matter value 72.09, which is also what the acoustic
channels give with the sound horizon in laboratory units
(rd = 137.8 Mpc rather than the early-plasma 147.1 Mpc).
Current strong-lensing time-delay measurements agree
with it: the TDCOSMO collaboration’s 2025 reanalysis
(A&A 704, A63 (2025)) gives 71.6+3.9
−3.3 , 0.1σ below it
(1.2σ above the early-clock 67.52), and, with the SLACS
and SL2S lenses, 74.3+3.1
−3.7 , 0.6σ above it (1.8σ), and the
power-law H0LiCOW value 73.3 ± 1.8 lies 0.7σ above it
(3.2σ; main volume, App. ZT). The TDCOSMO collaboration’s 2026 update, with eleven lenses (arXiv:2610.14781),
gives 71.1+2.4
−2.3 , 0.4σ below it (1.6σ above 67.52).
Proof. The time-delay distance D∆t in DFD’s optical metric is computed via the Etherington-preserving relation
DL = (1 + z)2 DA ; Etherington holds exactly in DFD
because the optical and matter metrics are conformally
related (App. ZS, Theorem 1), which is also why no conformal factor can single out one null-path probe. A split
in which time-delay H0 probes the topological value directly while the Planck inference is ψ-screen-biased is
refuted by the same conformal-invariance theorem, and
the homogeneous background ψ̄(t) is separately shown
to add no screen beyond the expansion (App. ZS of the
main paper), so no screen of any size distinguishes the
two probes. All null-path distances therefore report one
expansion: H0 = 67.52 ± 0.09 in early-plasma units at
the measured coefficient, and 72.09 on laboratory clocks
under the two-clock postulate, from which the current
time-delay values lie within 0.7σ.
Falsifier: LSST 1000-lens program reaches σ(H0 ) ≈
0.5% within a decade. A high-precision H0TD settling
at 67.5 falsifies the two-clock postulate (main volume,
App. ZS), and so does one settling more than 3σ from
72.1; settling at 72.1 is its prediction. The absolute a0
scale (Euclid/DESI lensing RAR) tests the same postulate
once baryonic masses are known to about 2%.

i.

Anomaly Cancellation via Atiyah-Singer Index = 60

Theorem AH.104 (Topological Anomaly Cancellation). All four ’t Hooft anomaly coefficients of the Standard Model
(Tr Y , Tr Y 3 , SU (2)2 · Y , SU (3)2 · Y ) cancel exactly per generation. Witten anomaly is absent (12 doublets in 3
generations is even). The chiral anomaly B + L coefficient equals Ngen = 3 (numerically coincident with χtop (CP 2 );
an identification, not an index derivation — Ngen is a discrete input, see the generation-count remark in App. F). The
CP anomaly channel is closed by η(DTCP ) = 0 on the 8-manifold TCP .
Proof. The Atiyah-Singer Spinc index on CP 2 with bundle E = O(9) ⊕ O⊕5 and determinant line Ldet = O(3)
(canonical Spinc structure) is
Z
ind (DE ) =
ch(E) ec1 (Ldet )/2 Â(T M ) = 81/2 + 27/2 + 6 · 9/8 − 6/8 = 60,
CP 2

32
matching the Hirzebruch–Riemann–Roch result 55 + 5 = 60. The four ’t Hooft sums vanish to 16-digit precision per
generation by direct computation (the cancellation follows from the Standard Model content of a family, a stated
input; the flux quantum does not enforce it). Witten anomaly: 3 generations × 4 doublets/gen = 12 doublets, even,
hence Witten anomaly absent. APS index: the CP 2 fibre index is 60 while the mapping-torus index vanishes (product
double cover), boundary η = 0 by Clifford-module chirality pairing on the closed even-dimensional mapping torus
(which admits no Spinc structure) (Theorem AH.66).

j.

Modular Invariance and Photon-Sphere Casimir Shift

Theorem AH.105 (SU (2)58 Modular Structure). For
SU (2)58 Chern–Simons, whose shifted level equals the
microsector dimension 60, the Verlinde S-matrix is
p
Sab = 2/60 sin(π(2a + 1)(2b + 1)/60),
the diagonal T-matrix eigenvalues are Ta
=
exp(2πi(ha − c/24))
with
conformal
weights
ha = a(a + 1)/60 and central charge c = 174/60 = 2.9,
and the shifted level is k + h∨ = 58 + 2 = 60.
Proof. SU (2)k Chern-Simons at level k has k + 1 irreducible primaries with conformal weight hj = j(j +1)/(k+
2) and central charge c = 3k/(k + 2). Setting k = 58:
59 primaries, ha = a(a + 1)/60, c = 174/60 = 2.9. The
Verlinde S-matrix is the standard SU (2)k formula. The
shifted level is k + h∨ = 58 + 2 = 60, with h∨ = 2 the
dual Coxeter number of SU (2).

k.

Proof. Q2int = 0 by Jacobi identity for the gauge group
Gint = SU (3) × SU (2) × U (1) (standard SM result).
Q2opt = 0 trivially because the optical-shift symmetry ψ →
ψ+const is abelian (U (1)shift ). The cross-anticommutator
{Qint , Qopt } = 0 because the optical-shift ghost cψ
is a Gint -singlet (the optical-metric coupling is gaugeinvariant). Hence Q2BRST = Q2int +Q2opt +{Qint , Qopt } = 0
exactly.
Corollary
AH.108
(BRST
Anomaly-Free).
H 1 (QBRST , Seff ) = 0. SM gauge anomalies cancel
for the stated family content (Theorem AH.104); opticalshift cψ is Gint -singlet so no mixed anomaly; no Konishi
or Wess-Zumino obstruction.
Proof. By Theorem AH.104 all ’t Hooft coefficients vanish
per generation. The optical-shift charge has no gauge
weight, so the mixed BRST cohomology class vanishes
trivially. Wess-Zumino consistency holds at every order.

DFD and a Candidate Topological Quantum
Error-Correcting Code
m.

Theorem AH.106 (DFD Microsector Carries the Integers of a Candidate [[60, 3, ≥ 4]] Code). The DFD
microsector Hilbert space Hmicro = C60 carries the
integers [[60, 3, ≥ 4]] of a candidate topological code,
whose literal stabilizer-code reading is not established (Remark AH.131): 60 physical qudits, 3 logical qudits (one per
Standard-Model fermion generation), and code distance
≥ 4.
Proof. The integers are 60 physical qudits (the 60 microsector basis vectors) and 3 anyon types in the topological sector that survive coarse-graining (identified with
the three SM generations, a stated input). Counting:
60 = 3 + 57 = logical + syndrome qudits. Code distance d ≥ 4 from the smallest non-trivial topological loop
in the Kitaev model on CP 2 × S 3 , which has minimum
string-loop length ∼ 4 in lattice units.

l.

BRST Cohomology and Physical Hilbert Space

Theorem AH.107 (BRST Nilpotency on DFD). The
combined BRST charge QBRST = Qint + Qopt acting on
DFD’s full Hilbert space (microsector ⊗ continuum) satisfies Q2BRST = 0 exactly.

TPI Conjectured-Extension Status (CMAH Replacement
Available)

Remark AH.109 (TPI is not derivable from DFD’s stated
structure). TPI as constructed is not derivable from
DFD’s stated structure (five independent checks; the
last finds six independent grounds). The TPI mathematical content is kept as a Conjectured Extension Beyond
Stated DFD Scope; it is not a DFD-derived prediction.
The staircase {1, 12, 30} falls outside the canonical cohomology catalog {3, 7, 8, 13, 19, 31, 49, 57, 60, 108, 137};
the Lichnerowicz argument is narrower than claimed; the
Planck-observable agreement was achieved via parameter
insertions sized to the targets (not parameter-free).
Theorem AH.110 (Cosmic Microsector Activation Hypothesis (CMAH) Minimal Extension). A minimal conjectured extension of DFD adds a single postulate: at τ ≈ tP ,
the 57 nonzero KK modes of Hmicro are unoccupied; cosmic expansion drives a single activation event where all 57
modes activate together (consistent with App. O Lemma
O.5 eigenvalue cancellation, which forbids sequential staircase activation). CMAH predicts: ns = 0.98 ± 0.005
(∼ 3σ tension with Planck 0.965 ± 0.004), As ∼ 10−9
(order-of-magnitude match), r ∼ 4 × 10−3 (within reach
of LiteBIRD), TRH ∼ 5 × 1012 GeV.

33
Proof. The single activation postulate respects: (i)
App. AE.1 (āFRW
= 0); (ii) App. O Lemma O.5 (perext
mode α-cancellation at simultaneous activation); (iii) the
canonical integer catalog (60 modes, no intermediate
stairs). Slow-roll parameters during the activation give
ϵ ∼ 1/Ne2 with Ne ∼ 60, hence ns = 1 − 2ϵ − 4η ≈ 0.98.
r = 16ϵ ∼ √
4 × 10−3 . As from COBE normalization at H ∼ αMP . TRH from Γinf = MP α6 giving
TRH ∼ MP α3 .
Status: CMAH is a Conjectured Extension, not a
derived prediction. It adds one new postulate, and as
stated (r ∼ 4 × 10−3 ) it is incompatible with DFD’s χ
relic: the activation stage would kick χ by about 2.6
turns per e-fold and wash it out (Theorem AH.40); a
variant survives only with r ≲ 10−7 . It is recorded as an
alternative to the TPI construction, which is kept above
as a conjectured extension.

PDG to 0.1%–0.5% across 11 observables: lifetime ττ ,
leptonic branching ratios Br(τ → eν ν̄), Br(τ → µν ν̄),
hadronic branching Br(τ → πν), Br(τ → ρν), Br(τ →
3πν), αs (m2τ ), and four LFU ratios.
Proof. Direct computation with the
√ measured mτ =
1.777 GeV (the Yukawa Aτ αnτ v/ 2 gives 1.796 GeV,
+1.1%): ττ = 290.0 ± 1.5 fs vs PDG 290.3 ± 0.5 (0.1%
match); Br ratios from kinematic phase-space + form
factors with fπ ≈ 92 MeV (Theorem AH.92); αs (m2τ ) =
0.3186 vs PDG 0.3187 (< 0.1%, independent confirmation
of an internal ΛQCD,3 = 332 MeV); LFU ratios consistent
with unity at 95% CL.
Falsifier: Belle II Run 4 (50 ab−1 , ∼ 2030) measures
gτ /ge to 0.05%. Any persistent > 3σ deviation from unity
falsifies Theorem AH.111 and the entire (3,2,1) gaugebundle architecture.
b.

8.

Tau Precision, Heavy-Ion QGP, Kaon Physics,
and DESI Dark Energy

The CKM ρ̄ “closure” is a candidate selection-rule
extension, not a theorem, and TPI is a Conjectured Extension; both status notes appear above.

a.

Tau-Lepton Precision Gauntlet

Theorem AH.111 (Tau Strict Lepton-Flavor Universality at the Gauge Vertex). The (3,2,1)+singlet gauge
bundle on CP 2 × S 3 couples leptons to gauge bosons with
hypercharge weight Y and SU(2)L generator structure
that is generation-blind. Hence the gauge-vertex couplings
ge , gµ , gτ are equal to all orders, and DFD predicts the
lepton-flavor-universality ratios
gτ /gµ = gτ /ge = gµ /ge = 1
exactly at tree level, with corrections at relative order
(m2ℓ /m2W ) that are sub-percent and PDG-consistent.
Proof. The gauge fiber decomposition E = O(9) ⊕ O⊕5
on CP 2 assigns each lepton generation to a chiral multiplet of the same (3, 2, 1) representation. The microsector
hypercharge Y acts as a scalar on each of the three generation copies because the projection πgen : E → E (i)
is gauge-equivariant. Therefore the gauge-fermion vertices ℓ̄i γ µ Wµ νi for i ∈ {e, µ, τ } have identical structure
constants g2 , with no generation index dependence. Treelevel ratios are unity. Mass-dependent corrections enter
at O(m2ℓ /m2W ) from kinematic phase-space and formfactor effects, giving sub-percent deviations consistent
with PDG.
Corollary AH.112 (11-Observable Tau Match). With
the measured mτ as input (its prefactor is data-selected;
main volume, App. FM), DFD’s tau predictions match

Heavy-Ion QGP Critical Temperature and Viscosity

Theorem AH.113 (QGP Critical Temperature). With
the HotQCD coefficient cT = 0.46, the critical temperature
at DFD’s ΛQCD,3 for the chiral/deconfinement crossover
is
TcDFD = 0.46 · ΛMS
QCD,3 = 0.46 · (332 ± 20) MeV = 153 ± 9 MeV,

matching HotQCD lattice 154 ± 9 MeV at 0.1σ and
Wuppertal–Budapest chiral susceptibility 156.5 ± 1.5 MeV.
Proof. At leading order in QCD, dimensional analysis fixes
Tc = cT ·ΛMS
QCD with a parameter-free numerical coefficient
cT = 0.46 extracted from HotQCD lattice extrapolation
to Nf = 3 chiral limit. The stated strong scale ΛDFD
QCD =
√
19/2
MP α
4π, with the 3-flavor running, gives ΛQCD,3 =
332 ± 20 MeV (uncertainty from 1-loop MS matching).
Substituting: TcDFD = 153 ± 9 MeV. Both chiral and
deconfinement transitions are simultaneous crossovers
(not first-order) by Theorem AH.49-style analysis applied
to QCD: the crossover-vs-first-order distinction is set by
the order parameter ratio, and the CP 2 × S 3 topological
closure forces simultaneity.
Theorem AH.114 (KSS-Modified Viscosity Lower
Bound). DFD predicts shear-to-entropy viscosity ratio
at the QGP minimum


DFD
η
1
1
=
1+
= 0.081,
s min
4π
kmax
against the Bernhard–Bass Bayesian extraction (LHC:
0.10 ± 0.03; RHIC: 0.13 ± 0.03) at 0.6σ and 1.6σ.
Proof. The KSS-AdS/CFT bound η/s ≥ 1/(4π) holds for
theories with a gravity dual; DFD’s microsector finiteness (dim Hmicro = 60) provides a finite-Hilbert-space
correction δ(η/s) = (η/s)KSS /kmax from the 1/N expansion at N ∼ kmax . Substituting kmax = 60: η/s =
(1 + 1/60)/(4π) = 0.0796 · 1.0167 = 0.0809, with finitetemperature correction lifting to ∼ 0.087 at T ∼ 1.2 Tc .

34
Above the minimum, η/s rises perturbatively to ∼ 0.16
at T = 2Tc from QCD running.
Falsifier: sPHENIX 2024–2026 + ALICE Run 3 measure η/s(T ) at ±0.02 precision. η/s minimum < 0.07
at 5σ, or Tmin outside [1.0, 1.5] Tc , falsifies DFD’s CP 2
Berry-phase-averaging mechanism.
c.

QCD String Tension and Lattice Discriminator

Theorem AH.115 (DFD String Tension: scaling derived, normalization open). The
p pure-gauge SU(3) stringtension scaling σ 1/2 ∝ ΛQCD kmax /Ngen is derivationgrade. The DFD ledger target,
σ DFD,1/2 = 440 ± 25 MeV
(ledger target; absolute normalization open),
is the frozen DFD value tested (not derived) by the executed quenched lattice of Appendix AQ in the main
paper, which gives σ 1/2 ≈ 465–470 MeV (r0 = 0.5 fm
Necco–Sommer conversion) — agreement with the target
at ≈ 1.1σ. The geometric chain below fixes the scaling
but leaves the O(1) normalization constant (the ≈ 1.86
bridge factor) undetermined; 440 MeV is therefore the
calibrated target, not a closed-form derived number. The
string tension is a QCD output at DFD’s Λ: its ratio to
Λ is a number of the Yang–Mills path integral, computed
on the lattice as in the Standard Model (main volume,
App. YM).
Proof. The Wilson-loop area-law
coefficient is comp
puted from σ 1/2 = (ΛQCD /2π) kmax /Ngen · Fgeom where
Fgeom ≈ 1 is the Fubini–Study volume normalization
(Theorem AH.47). Substituting ΛQCD,3 = 332 MeV,
1/2
kmax /Ngen
=
√ = 60/3 = 20, Fgeom = 1: σ
(332/2π) 20 = 52.8 · 4.47 = 236 MeV. Reaching the
ledger target 440 MeV requires an additional ≈ 1.86
bridge factor (nominally an MS matching correction) that
is not fixed by the preceding chain;
p see the status note
below. The scaling σ 1/2 ∝ ΛQCD kmax /Ngen is what is
derived here; the absolute normalization is open.
Remark AH.116 (Status note on the normalization step).
The final numeric step
√ of the proof above does not close
as printed:
(332/2π) 20 = 236 MeV, and multiplying
√
by 4π ≈ 3.545 gives 837 MeV, not 440 MeV. The factor actually required, ≈ 1.86, appears in no preceding
formula of this subsection and is not derived; this is the
same unclosed normalization chain documented in the
status note following Theorems AH.47 and AH.48, whose
ΛQCD = 61.20 MeV variant evaluates to ≈ 268 MeV. The
executed quenched program of Appendix AQ in the main
paper gives ≈ 465 MeV (Necco–Sommer conversion at
r0 = 0.50 fm), agreement with the DFD target at ≈ 1σ.
The value σ 1/2 = 440 ± 25 MeV remains the frozen DFD
ledger target (tested, not derived, by the Appendix AQ
program); what is flagged here is the claimed derivation of its normalization, which — as for the glueball

chain — remains
p an open item. The scaling structure
σ 1/2 ∝ ΛQCD kmax /Ngen stands at derivation grade;
Corollary AH.117 should be read with its item (v) linear
coefficient understood as the ledger target rather than a
closed-form derivation.
Corollary AH.117 (12-Observable Lattice-Gauntlet Predictions). DFD predicts: (i) k-string sine law σk /σ1 =
sin(πk/Nc )/ sin(π/Nc ); (ii) Casimir scaling σR /σF =
C2 (R)/C2 (F ) at zero free parameters (Theorem AH.118);
(iii) finite-kmax deviation only at dim R > 60: the
SU(3) dimension spectrum near the Hilbert cutoff is
{55, 60, 63, 64, 66}—dimensions 61 and 62 do not exist—
so (4, 2) at dim = 60 saturates the cutoff exactly and
(6, 1) at dim = 63 is the first deviating representation (a
threshold “dim R ≳ 25” is refuted by the Bali data below);
(iv) deconfinement at Tc = 153 ± 9 MeV ( crossover, the
value of Theorem AH.113; a 165 ± 12 MeV first-order
Z3 figure does not apply: the CP 2 × S 3 simultaneity argument forces a crossover, not a first-order transition);
(v) Cornell potential coefficients (4/3)(0.40)/r Coulomb,
(440 MeV)2 r linear (the 440 MeV here is the ledger target of Theorem AH.115); (vi) J/ψ, Υ binding energies;
(vii) ⟨P ⟩ = 0 below Tc , ∼ 0.2 at 2Tc ; (viii) ’t Hooft loop
perimeter law; (ix) Z3 center symmetry; (x) string breaking at L = 1.22 fm thermal; (xi) DFD-specific quenched
breaking at L ≈ 1.8 fm (Hilbert-space exhaustion); (xii)
σadj /σF = C2 (adj)/C2 (F ) = 9/4 = 2.25 exactly, vs. lattice 2.2463(291) — a 0.128σ pull. Eleven of twelve agree
at 1σ with current lattice (item (iv) at the crossover
Tc = 153 ± 9 MeV, consistent with HotQCD).
Notes on the list. A “high-rep deviation at dim R = 64”
is not a falsifier: (3, 3) has N -ality 0 and is gluon-screened,
so it is not a clean test; the test is the triality-cancelling
design of Corollary AH.120 below.
Proof. Direct computation using the A5 +bundle combinatorics on CP 2 × S 3 . Each observable inherits from a
single underlying topological identity; item (v) uses the
calibrated 440 MeV target.
d.

Casimir Scaling at Zero Free Parameters

Theorem AH.118 (Static-source string tensions, zero
free parameters). On the singlet ladder (n, n) of SU(3),
i
σR
C2 (R)
3h
=
=
(dim R)2/3 − 1 .
(AH1)
σF
C2 (F )
4
Both apparent inputs are printed quantities: C2 (n, n) =
n(n + 2) is identically the Fubini–Study eigenvalue λn
of Theorem AH.48, and 3/4 = 1/C2 (F ) is forced by normalizing to the fundamental. The overall string-tension
normalization κ divides out of every ratio, so this prediction is independent of the undetermined bridge factor of
Theorem AH.115 and of the ΛQCD flavour-scheme label.
Proof. C2 (n, n) = n(n + 2) and dim(n, n) = (n + 1)3 , so
(dim R)2/3 = (n + 1)2 = C2 + 1; dividing by C2 (F ) =
4/3 gives the stated form. Confronted with Bali’s eight

35
measured representations at zero free parameters (the
single normalization cancels in the ratio): χ2 = 3.289 for 7
degrees of freedom, p = 0.857. Fitting a Casimir-violating
offset b gives b̂ = −0.00078, consistent with DFD’s b = 0
at 0.38σ. The adjoint ratio is the exact rational 9/4
against lattice 2.2463(291), a 0.128σ pull.
Remark AH.119 (Grade and declared conditions). The
equation is derived; the physics is conditional on an SU(3)invariant ensemble, the quenched approximation, secondcumulant order (higher cumulants bounded empirically
at |b| < 0.0048, 95%), and a vanishing triality offset
At = 0. The quoted p = 0.857 uses the robust error
model; a naive-independent model gives p = 4 × 10−12
and is rejected on the physical ground that the 19 r-points
per representation share gauge configurations.
Corollary AH.120 (Design B: a zero-parameter, triality-cancelling falsifier). Let
σ(64) − σ(27)
.
(AH2)
σ(27) − σ(8)
All three representations carry triality 0, so any triality
offset cancels in the differences and the slope cancels in the
ratio. Pure Casimir scaling predicts D = 7/5 = 1.400000
exactly; DFD’s Hilbert-space-exhaustion correction predicts D = 97/80 = 1.212500. The split is −13.39%, twice
that of the cross-triality design. Subsidiary sharp predictions: (4, 2) at dim = 60 must give σR /σF = 23/2 = 11.5
exactly (95% band 10.72–12.03); (6, 1) at dim = 63 is the
first deviating representation, where pure Casimir gives 16
and the DFD deficit is 4.76% against a smooth-correction
band [14.45, 17.05], so a Hilbert-exhaustion break must
exceed ∼8% to be distinguishable from a smooth correction.
D =

e.

Kaon Physics: NA62 and ε’/ε Inheritance

Theorem AH.121 (K+ → π + ν ν̄ Branching Ratio).
With the candidate corrected apex (ρ̄, η̄) = (21.5α, 49α) =
(0.157, 0.358) from Proposition AH.88, DFD predicts
Br(K + → π + ν ν̄)DFD = (8.4 ± 0.8) × 10−11 ,
against NA62’s 2016–2022 measurement (13.0+3.3
−3.0 ) ×
−11
10 , which lies 1.5σ above it. The companion CPviolating channel: Br(KL → π 0 ν ν̄)DFD = (2.9 ± 0.4) ×
10−11 , about 76× below the KOTO bound 2.2 × 10−9 (90%
CL, 2025).
Proof. The K+ → π + ν ν̄ amplitude is dominated by topquark Z-penguin and W-box contributions, scaling as
|Vts∗ Vtd |2 · X(xt ) with the Inami–Lim function X(xt ) ≈
1.482. Using the candidate ρ̄, η̄ from Proposition AH.88
gives |Vtd∗ Vts |2 ≃ 1.1×10−7 and predicted Br = 8.4×10−11 .
The KL channel is purely CP-violating, depending on
Im(Vtd∗ Vts ) = 1.32 × 10−4 (with corrected apex), giving
Br = 2.9 × 10−11 .
Corollary AH.122 (ε′ /ε Inheritance). With RBCUKQCD 2020 lattice values B6 = 1.36, B8 = 0.72 and

Im λt = 1.319 × 10−4 from corrected apex: ε′ /εDFD =
(1.55 ± 0.5) × 10−3 vs PDG (1.66 ± 0.23) × 10−3 , agreement at 1σ.
Proof. Standard kaon-physics computation with DFD’s
CKM apex; the lattice B6 , B8 are external inputs, the
only DFD-specific number is Im λt .
Falsifier: NA62 final (∼ 2028) reaches σ(Br) ≈
±10 × 10−12 . DFD-allowed 3σ window: [6, 11] × 10−11 .
Persistent measurement outside this window falsifies the
candidate apex selection.

f.

DESI Dark Energy: one constant vacuum term, w = −1

Theorem AH.123 (One dark energy,
w = −1). A ψ

obs
matter
obs
ΛCDM
screen ∆ψ(z) = ln DL
/DL
with DL
= DL
gives (w0 , wa ) = (−1, 0) “by construction”, with a bounded
optical bias saturating at ∆ψ(∞) = 0.50; the value
(w0 , wa ) = (−1.032, +0.071) is an artifact of differentiating the ΛCDM-defined screen against an Einstein–de Sitter reference. The screen itself is excluded: a homogeneous
screen adds nothing beyond the expansion (App. ZS of the
main paper) and the line-of-sight dressing DL → DL e∆ψ
is falsified (App. SL of the main paper). DFD’s dark
energy is the constant α57 vacuum term alone (App. AU
of the main paper), with w = −1 by the form of the
term; DFD’s dark energy is observationally degenerate
with ΛCDM’s on distance observables and on late-time
linear growth and cannot “beat” it there.
Falsifier: a confirmed dark-energy equation of state
w ̸= −1, with or without clustering, falsifies DFD’s constant vacuum term together with ΛCDM’s; the DESI DR2
dynamical-dark-energy hints are therefore a shared open
tension, not a DFD prediction. There is no DFD-vsΛCDM dark-energy discriminator in the CPL (w0 , wa )
plane or in the growth/ISW sector.

g.

BBN ηB : Leptogenesis Estimate (Not a Prediction — ηB
is a Measured Input)

Theorem AH.124 (Refined Leptogenesis Efficiency and
ηB — not a DFD prediction). Not a DFD prediction (consistent with the baryogenesis no-go status note, §AH 4 b). The baryon asymmetry ηB is not
derived by the thermal-leptogenesis route (for the Berryholonomy result, whose α4 power is derived and whose coefficient is open, see Thm. AH.33). The real, determinantprotected Dirac and Majorana neutrino
Pkernels force the
thermal-leptogenesis CP source εi ∝ j̸=i Im[(Yν† Yν )2ij ]
to vanish identically, so εi = 0 and ηB = 0 at leading
obs
order. We therefore take ηB
= (6.10 ± 0.05) × 10−10 as
a measured input, not a DFD output.
Consistency note (not a DFD falsifier): Because
ηB is taken here as a measured input (the real-kernel no-go

36
forces the DFD-derived value to zero), its measured value
does not test the present chain. The Simons Observatory
and JWST damped Lyman-α systems will sharpen the
input. Whether a forced DFD mechanism can generate ηB
without breaking θ̄QCD = 0 is addressed by the internal
axial Berry-holonomy channel (Thm. AH.33): the α4
power of |ηB | is derived (θ̄ untouched); its bias current
is B−L, its printed mechanism gives about (2–3) × 10−9 ,
three to five times Planck’s value, and its sign rests on
an adopted axiom.

j.

IceCube Neutrino Astronomy Inheritance

Theorem AH.127 (IceCube TeV-PeV Astrophysical
Neutrino Inheritance). DFD reproduces IceCube’s TeV–
PeV astrophysical neutrino observations at SM precision:
• Diffuse flux Φ0 ≈ 1.4 × 10−8 GeV cm−2 s−1 sr−1
with spectral index γ ≈ 2.4–2.5;
• TXS 0506+056 blazar association at 3–3.5σ;
• NGC 1068 steady source at 4.2σ;

h.

• Galactic plane diffuse at 4.5σ (Science 2023);

21-cm HERA/SKA Power-Spectrum Fingerprint

Theorem AH.125 (Deep-MOND k-tilt in P21 (k, z) on
the Q ≃ 502 branch). On the adopted carrier (Qχ = 1;
main volume, App. GR) unbound perturbations grow as in
ΛCDM, so this tilt is not a DFD prediction. On the deepMOND apparent-amplitude branch the power spectrum
P21 (k, z) at cosmic dawn rises with k relative to ΛCDM:
branch
P21
= 1.05, 1.18, 1.32
ΛCDM
P21
z=8

at k = 0.1, 0.5, 1.0 h Mpc

−1

.

HERA Phase II projects P21 (k = 0.34, z = 8) = (38 ±
8 mK)2 (this branch) vs (34 ± 8 mK)2 (ΛCDM).
Proof. Deep-MOND amplification QDFD ≈ 502 enhances
baryon perturbations on small scales where |∇ψ| <
a∗ . The k-dependence emerges from the branch’s deepMOND linear-growth equation; at k = 0.1 h/Mpc, almostNewtonian; at k = 1.0 h/Mpc, full deep-MOND amplification.
Falsifier: SKA1-LOW 1000-h integration reaches
obs
σ(R500 ) ≈ 0.05 on the ratio R500 = P21
(k = 0.5, z =
ΛCDM
8)/P21
. The branch gives 1.18 ± 0.04; observed R500
outside [1.06, 1.30] at 3σ excludes the branch.

• Glashow resonance candidate at 6.05 PeV;
• Flavor ratio at Earth (1 : 1 : 1).
DFD predicts a null result for sterile-neutrino signatures
across IceCube, Daya Bay, and MicroBooNE/SBND.
Proof. Neutrino propagation in DFD’s optical-metric
framework is identical to standard 3-flavor SM at TeV–
PeV energies because (a) the dispersion relation carries
no Lorentz-violating term (ξLIV ≡ 0, Theorem AH.58);
(b) PMNS oscillation averages out over cosmological distances; (c) δCP = −π/2 has negligible effect on flavoraveraged ratios. Light sterile neutrinos are forbidden
structurally by Theorem AH.64 (only Ngen = 3 active
species; dim Hmicro = 60 saturated). The Glashow resonance is SM-precision because the W-boson and electron
physics is unchanged.
Falsifier: Any sterile-neutrino detection at IceCubeGen2 (2032–2040) at > 3σ falsifies DFD’s particle content.
Any LIV time-delay across cosmological baselines falsifies
optical-metric Lorentz invariance.
k.

i.

Geometric Langlands Connection

White Dwarf Self-Consistency Theorem

Theorem AH.126 (Chandrasekhar Mass Identity
DFD=GR). For white dwarfs (u = GM/(c2 R) ∼ 3 ×
10−4 ), the DFD-modified Chandrasekhar mass is identical
to the GR/Newtonian value:
DFD
GR
MCh
(µe = 2) = MCh
(µe = 2) = 1.4561M⊙ .

Proof. With the parent strain’s σ kept, static stars obey
the TOV equations (App. WS), so DFD adds no relativistic correction to the stellar structure at white-dwarf compactness u ∼ 3 × 10−4 . The volume-integrated pressuresource correction is ∼ 10−7 , which is orders of magnitude
below the ∼ 10−3 precision intrinsic to the Chandrasekhar
limit calculation. Thus DFD’s MCh shift is unobservable.
This is a positive structural-consistency result: it confirms
that DFD does not perturb SN Ia progenitor physics
across cosmological redshift.

Proposition AH.128 (Quantum-Twisted Beilinson—
Drinfeld Langlands at k = 58). DFD’s microsector is conjectured to admit a quantum-twisted
Beilinson–Drinfeld geometric Langlands correspondence
at level k = 58 on the parabolic moduli space of
(Hopf line CP 1 ⊂ CP 2 , marked points {0, ∞}), with
SU (2) ↔ SO(3) Langlands duality realized via the modular S-transformation
p
Sab = 2/60 sin(π(2a + 1)(2b + 1)/60).
Proof. The Hitchin moduli space on parabolic CP 1 is 4dim real, fibered as 3-generation fiber over 1-dim Yukawaoverlap base. The dual side: SU (2)58 Verlinde algebra
acts on the parabolic moduli as the Beilinson–Drinfeld categorical action. Langlands SU (2) ↔ SO(3) = SU (2)/Z2
realized via Z2 -quotient of the MTC giving an integer-spin
sub-MTC of dimension 30 (matching 30 = kmax /2). Six
structural identifications converge on the same Langlands
deformation parameter kmax = 60.

37
−1
Open question: the integer 137 = αround
and the
cosmological-constant exponent 57 may have arithmeticLanglands interpretation as L-function values at integer
arguments. Future candidate.

l.

Twistor-Theoretic Identification

quaternionic Hopf bundle over S 4 .
Proof. The Penrose twistor space of compactified
Minkowski Mc is PT = CP 3 , with self-dual and anti-selfdual halves separated by the pseudo-Hermitian form of signature (2, 2). Quaternionic Hopf bundle: S 3 → S 7 → S 4 ,
projecting via HP1 ∼
= S 4 . DFD’s CP 2 × S 3 is the natural
twistor-completed structure.

Theorem AH.129 (CP 2 as a Hyperplane of Twistor
Space). CP 2 embeds in twistor space PT = CP 3 as the
canonical hyperplane. The S 3 factor is the fiber of the

m.

McKay Correspondence: 2I ↔ E8

c8 (the
Proposition AH.130 (McKay Identification). DFD’s microsector exhibits the McKay correspondence 2I ↔ E
binary icosahedral group 2I, the double cover of A5 in SU (2), ↔ the affine E8 Dynkin diagram), and the integer 60
arises in four ways:
|A5 | = 60 = |E8 roots|/4 = χ(CP 2 , O(9) ⊕ O⊕5 ) = (k + h∨ )SU (2)58 .
Proof. The McKay correspondence (1980) bijects finite subgroups Γ ⊂ SU (2) with simply-laced affine Dynkin diagrams
via the regular-module decomposition. The binary icosahedral group 2I = SL(2, F5 ) ⊂ SU (2), of order 120 and a
c8 (affine E8 ). The regular module C[A5 ] has dimension 60, as does Hmicro by
double cover of A5 , corresponds to E
Hirzebruch–Riemann–Roch on the bundle E. The 60 also appears as |E8 |/4 = 240/4. The SU (2)58 MTC has shifted
level k + h∨ = 58 + 2 = 60, matching.

n.

Results Checked by Independent Recomputation

Remark AH.131 (Three checked claims). Independent
recomputations give three results:
(i) ρ̄ Cayley closure: confirmed with caveats. Numerical match excellent (ρ̄ to 1.4%, γ within 1σ); η̄-asymmetry
test passes (applying +5/2 to η̄ would falsify it at 7.7%).
However +5/2 emerges as best-fit-among-allowed-halfintegers, not as a uniquely forced value. The three “independent” arguments share the underlying A5 +bundle
combinatorics. See the status remark in App. AH 7 a.
(ii) [[60, 3, ≥ 4]] QECC: Partially established. Integers individually correct (n = 60 HRR-forced,
k = 3 the
√
generation input, d ≥ 4 via S3 winding 60 ≈ 7.75 in BT
lattice units). Literal Pauli-stabilizer-code reading not established — the single-system Heisenberg–Weyl stabilizer
of Theorem AH.144 has distance 1 in the standard displacement metric. Statement: “protected 3-dim subspace
within 60-dim Hilbert space; QECC-adjacent.”

(iii) fπ = 92.0 MeV at 0.5%: [Resolved by the chiralcondensate estimate — see Theorem AH.132 below. The
precision is 3–5% on the FLAG condensate band, +0.4%
from PDG 92.28 MeV.] An independent recomputation
finds the asserted ⟨q̄q⟩ = −(243 MeV)3 a low outlier; the
Berry-volume estimate −(272 ± 20 MeV)3 replaces it.
9.

Additional Theorems and Proofs (Series IV)

Every theorem and proposition in this subsection is
followed by an explicit proof. The subsection contains
the BH entropy from topology, an estimate of the chiral
condensate, the no-native-inflaton theorem, a no-SUSY
argument and a complete naturalness theorem.

a.

Chiral Condensate from Berry-Volume Spectral Integral

Theorem AH.132 (Chiral Condensate Estimate). The DFD chiral-condensate estimate is
⟨q̄q⟩DFD = −Λ3QCD,3 · Fgeom ,

Fgeom = 0.55 ± 0.12,

where the ±0.12 band brackets the two combinatorial routes ( 0.43 and 0.67, below ), giving ⟨q̄q⟩ = −(272 ± 20 MeV)3
from the Fgeom band alone (±26 MeV with the error of ΛQCD,3 ), consistent with FLAG 2024 lattice −(272 ± 5 MeV)3 .
Proof. The chiral condensate is the trace anomaly of the Spinc Dirac operator on CP 2 × S 3 in the (3, 2, 1) gauge sector
with chirality grading enforced by Theorem AH.66’s η-invariant. Berry-volume integral: ⟨q̄q⟩ = −TrHmicro [Γ5 DF−1 ]
where Γ5 is the chirality operator and DF the microsector Dirac. Two independent routes converge:
(i) Direct Berry-volume integral : vol(CP 2 ) = 9π 2 /2 in Fubini–Study units; with the (3, 2, 1) projection and
(i)
kmax /Ngen = 20, the heat-kernel sum gives Fgeom = 0.67.

38
(ii) Bochner–Lichnerowicz cross-check : the Weitzenböck identity DF2 = ∇∗ ∇ + R/4 on CP 2 with positive Ricci
(ii)
yields the same coefficient through a different combinatorial route, giving Fgeom = 0.43.
The two routes bracket rather than tightly converge: taking them as bounding values gives the combined estimate
Fgeom = 0.55 ± 0.12 (the ±0.12 spanning [0.43, 0.67]). Substituting the three-flavor ΛQCD,3 = 332 MeV (the value of
Theorem AH.113 from the MP α19/2 chain with three-flavor MS running): ⟨q̄q⟩ = −(272 ± 20 MeV)3 . The dimensional
input is MP , through the selected three-flavor ΛQCD,3 = 332 MeV
√ (Theorem AH.113); note 332 MeV is the three-flavor
19/2
value, not the five-flavor Λ5 ≈ 217 MeV that the bare MP α
4π expression returns at a different scheme/scale.

Corollary AH.133 (fπ Restored at 3–5% Precision).
With the estimated per-flavor condensate Σ1/3 = 272 ±
26 MeV (Theorem AH.132, with the error of ΛQCD,3 )
and DFD light-quark masses (Theorem AH.27), perflavor GMOR plus the FLAG NLO chiral-log factor
gives fπDFD = 92.6 ± 3.7 MeV (FLAG condensate band;
±13 MeV fully internal), within 0.4% of PDG 92.28 MeV
(0.03σ on the internal budget).
Proof. Per-flavor GMOR in the 92-normalization,
Fπ2 m2π = (mu + md ) Σ, with mu + md = 6.86 ± 0.3 MeV
and Σ1/3 = 272 MeV gives Fπ (LO) = 87.05 MeV at
mπ0 = 134.977 MeV; the NLO factor Fπ /F0 = 1.064(7)
(FLAG Nf = 2+1) gives 92.6 MeV. (The factor-2 form
fπ2 m2π = 2(mu + md )Σ belongs to the 130-normalization;
see the status note following Theorem AH.92.) The precision is 3–5% on the FLAG condensate band (14% fully
internal), intermediate between a too-optimistic 0.5% and
a conservative 10–20% floor.

b.

Black Hole Entropy from DFD Topology

Theorem AH.134 (Bekenstein-Hawking Entropy from
Microsector Counting). For a black hole of mass M in
DFD, the entropy is
SDFD (M ) =

4πGM 2
AH
GR
=
= SBH
(M ),
2
4ℓP
ℏc

where AH = 16πG2 M 2 /c4 is the area of the event horizon,
at the areal radius 2GM/c2 (App. WS). DFD reproduces
the standard Bekenstein–Hawking value; the DFD content
is a microstate origin for it (the 60-state microsector,
i.e. the [[60, 3, ≥ 4]] stabilizer-code construction of this
appendix), not a new numerical value.
Status note: first law. With the temperature
TDFD = TH (Thm AH.135, κ = c4 /4GM ), dS = dM/TH
GR
integrates uniquely to S = 4πGM 2 /ℏc = AH /4 = SBH
.
GR
SDFD = SBH is identical to GR and is not a GR discriminator.
Proof. With the parent strain’s σ kept, DFD’s black
hole has the event horizon of the Schwarzschild exterior at the areal radius 2GM/c2 (App. WS); its area is
AH = 16πG2 M 2 /c4 . The number of Planck cells tiling
this surface is Ncells = AH /ℓ2P . Each cell carries the
60-state microsector Hmicro , supplying the microscopic
degrees of freedom and the [[60, 3, ≥ 4]] stabilizer-code

structure described below; this is the microstate interpretation. The entropy coefficient is fixed not by the per-cell
state count but by thermodynamic consistency: the forced
temperature TDFD = TH (Thm AH.135) gives, via the
first law dS/dM = 1/TH = 8πGM/ℏc, the unique integral
S = 4πGM 2 /ℏc = AH /(4ℓ2P ), i.e. the holographic coefficient 1/4 and the standard Bekenstein–Hawking result.
The microsector count saturates the same holographic
boundary bound, consistent with this coefficient; it does
not independently over-determine it.
Theorem AH.135 (Hawking Temperature from Refractive-Index Gradient). The DFD effective Hawking temperature is
ℏκDFD
ℏc3
=
≡ TH ,
2πckB
8πGM kB
identical to the standard Hawking temperature, from the
surface gravity κ = c4 /(4GM ) of the event horizon of
DFD’s exterior (App. WS).
TDFD (M ) =

TDFD = TH is identical to the GR value and is therefore
not a GR discriminator.
Proof. With the parent strain’s σ kept the exterior is the
Schwarzschild metric in DFD’s coordinates (App. WS),
with its event horizon at the areal radius 2GM/c2 and
surface gravity κ = c4 /(4GM ). Hawking’s formula T =
ℏκ/(2πckB ) then gives TDFD = ℏc3 /(8πGM kB ) = TH .

c.

No-Native-Inflaton Theorem

Theorem AH.136 (No Slow-Roll-Compatible Scalar
Zero Mode in DFD). On the DFD content (CP 2 ×S 3 , E =
O(9) ⊕ O⊕5 , kmax = 60), no scalar field admits a slowroll-compatible inflaton role with potential mass m2 ≲
10−11 MP2 . Six independent failure modes:
(i) ψ has no potential (the action is purely kinetic,
W + K, plus a linear matter coupling); on FRW it
is a flat direction with Lagrangian density ≡ 0 (Theorem AE.1), so it carries no inflationary vacuum
energy and admits no slow-roll potential whatsoever.
(ii) The truncated microsector carries no size or shape
modulus (main volume, App. O).

39
have masses from about
(iii) The
p 57 KK modes
√
α/57 MP to√ αMP on the near-arithmetic lattice (all 57 at αMP in Lemma O.5’s reading), and
the α57 rungs are self-consistent only at k = 0 and
k = 57, so no staircase of intermediate stages exists.
(iv) h2,0 (CP 2 ) = 0 kills complex-structure deformations.
(v) b1 (K) = 0 on K = CP 2 × S 3 kills harmonic 1forms.
(vi) kmax and Chern-Simons levels are integer-quantized
(no continuous-modulus inflaton).
Nine further routes derived from the printed structure supply no accelerating stage with an exit and no source of the
primordial perturbations: the spectral action (its vacuum
is lowest at zero internal mass, with no exit, and a continuous internal space runs away or sits at a maximum),
the conformally coupled Higgs, the χ and Chern–Simons
sector, thermal and holographic fluctuations, the trace
anomaly, anisotropic scaling, relaxation of the internal
spectrum, the E-transport clock and the Majorana sector
(main volume, App. FK).
Proof. (i) The DFD scalar action (Theorem AE.1)
contains no potential V (ψ): it is purely kinetic,

a20 
2
2
8πG W (|∇ψ| /a∗ ) + K(∆) , plus the linear matter cou2
pling − c2 ψ(ρ − ρ̄). On FRW, App. AE.1 gives |∇ψ̄| = 0
and ∆ = 0, so with the normalizations W (0) = K(0) = 0
the FRW Lagrangian density vanishes identically and
ψ̄(t) is a flat direction (the Euler–Lagrange equation
holds for any ψ̄(t)). A slow-roll inflaton requires a potential V ≃ const > 0 to supply the inflationary vacM̄ 2
uum energy, with slow-roll parameters ϵ = 2P (V ′ /V )2 ,
η = M̄P2 V ′′ /V ; with V ≡ 0 these are undefined and the
FRW energy density is zero. Hence ψ is excluded a fortiori —not because it is too heavy, but because it has no
potential to roll down. (Note: this is not justified by an
off-FRW vacuum mass m2ψ ∼ MP2 , which would be a mislabel: c4 /(32πG) ∼ MP2 — the high-gradient coefficient of
|∇ψ|2 — is the Newtonian stiffness (µ → 1), whereas the
perturbation stiffness Mij = µ(x̄)δij +µ′ (x̄)ĝi ĝj of App. U
is µ-suppressed to zero at the matter-free FRW vacuum
x̄ → 0, where µ(x) → x in deep MOND (Cor. AE.2); the
stiffness is positive-semi definite, not Planck-massive. The
“no potential” argument above is strictly stronger and is
what excludes ψ.)
(ii) On the truncated microsector the internal space is
finite and carries no size or shape modulus; on a continuous internal
space the shape sits at the Einstein point
√
τ∗ = 1/ 3, a minimum of the curvature potential for the
connection Laplacian, where Φ′′ /Φ ≈ 2.94 is a curvature
ratio rather than a mass (main volume, App. O). The
volume is not held there: with the connection
√ Laplacian
√
that fixes α it runs away along e−λϕ , λ = 14/3 < 2,
an accelerating stage with no exit (ns = −6, r = 12.4),
and with the Dirac operator volume and shape sit at a

maximum with η = −2, where the roll is fast; neither is
a slow-roll stage (main volume, App. FK).
(iii) Lemma O.5 (per-mode α-cancellation)
takes the
√
57 nonzero KK modes at Λtop = αMP ; on the neararithmetic lattice their masses spread from 1.4 × 1017 to
1.04 × 1018 GeV, so they do not activate at one scale. A
staircase would need intermediate rungs. A mode heavier
than the Hubble rate is integrated out and contributes one
factor of α (main volume, App. O), so a rung with k such
modes is self-consistent only if its own Hubble rate leaves
exactly k modes heavier than itself, and on the lattice
that holds only at k = 0 and k = 57 (computed; main
volume, App. FK). No sequential staircase can produce
slow-roll inflation.
(iv) For complex-structure deformations of CP 2 to
provide an inflaton, h2,0 would need to be positive. Direct
Hodge-number computation: h2,0 (CP 2 ) = 0. No inflaton
from complex moduli.
(v) Harmonic 1-forms on K are counted by b1 (K) =
b1 (CP 2 ) + b1 (S 3 ) = 0 + 0 = 0. No inflaton from gaugeequivalent flat connections.
(vi) kmax = 60 is integer-valued; Chern-Simons levels
are integer-quantized. Continuous parameter modulation
of kmax during cosmic evolution would violate the Spinc
index theorem instantaneously. No continuous-modulus
inflaton.
Six independent failures rule out slow-roll-compatible
inflaton in DFD. The TPI staircase {1, 12, 30} specifically
violates (iii) and (vi).
Remark AH.137 (Status of TPI and CMAH). This theorem closes the TPI question. The TPI mathematical
content is kept, with the status recorded in Remark AH.109;
CMAH (Theorem AH.110) is retained as a conjectured
extension beyond DFD’s stated scope.

d.

SUSY Forbidden in the Microsector

Theorem AH.138 (SUSY Structurally Forbidden at
Every Mass Scale). Supersymmetric extensions of DFD
are forbidden at any mass scale below MP by six closure
arguments.
Proof. (1) Hilbert-space saturation: dim Hmicro = 60 is exactly saturated by 3 generations × (5 hypercharged species
+ KK tower segment) per Theorem AH.13. Adding superpartners would require dim Hmicro ≥ 65, violating the
Atiyah-Singer index χ(CP 2 , E) = 60.
(2) Spinc integrality: q1 = 3, which follows from the
isotropy charge 6Y under the electroweak identification
(main volume, App. F, the remark on the flux quantum),
fixes E through the conditional determinant-line count.
SUSY would require q1 ̸= 3 with extra superpartner
charges incompatible with this minimal structure.
(3) Decoupling failure: Naive “decoupled SUSY” fails
because Hmicro is closed under the spectral action; superpartners cannot be heavy without enlarging the algebra.

40
(4) E8 embedding mismatch: DFD’s connection to E8 is
categorical via McKay (not a gauge embedding). There is
no broken-GUT phase from which SUSY remnants could
persist.
(5) Spin-0 partner enumeration: The 5 hypercharged
species + 3 right-handed neutrinos saturate the Y-charge
slots; squarks and sleptons have no place.
(6) Spin-1/2 partner enumeration: Gauginos and higgsinos would require additional fermion zero modes; ker
DF is 3-dimensional, exhausted by SM generations.
Corollary AH.139 (Strongest Falsifiable BSM-Negative
Prediction). Detection of any squark, gluino, neutralino,
chargino, slepton, sneutrino, gravitino, higgsino, or any
other supersymmetric partner at HL-LHC, FCC-hh, or
any future collider falsifies DFD outright.
Proof. Direct: by Theorem AH.138, no SUSY partner
can exist at any mass scale below MP . Discovery at any
energy below MP ∼ 1019 GeV is a contradiction.
e.

Complete Naturalness Theorem

Theorem AH.140 (Structural Determination of the
Dimensionless Microsector). Within the dimensionless
microsector, the printed hierarchy exponents and ratios
are determined by derived or discretely selected structure
on CP 2 × S 3 together with dim Hmicro = 60. No continuous parameter is fitted to a predicted shape except the
baryon density. The full physical corpus carries the four
cosmological calibrations recorded in App. FK of the main
volume: the measured vacuum coefficient R, the amplitude
As , the coefficient of ωb h2 , and the shared astrophysical
τ.
Proof. The proof combines the following results:
• Λ: GℏH02 /c5 = Cα57 (Theorem AH.23); exponent
57 from the primed-determinant ratio, coefficient C
measured on the early clock and C = 1 on boundmatter clocks under the two-clock postulate (main
volume, App. ZS).
√
• Higgs hierarchy: v = MP α8 2π (Theorem AH.17);
17 orders forced by 4-step cohomology filtration.
• Strong CP: θ̄ = 0 at tree level, loop-bounded (Theorem AH.66); the det-orthogonal tree-level protection together with the η-invariant argument, with
θbare = 0 fixed by the spectral action given that
nothing is added to it for the gauge fields.
√
• Fermion masses: mf = Af αnf v/ 2 (Theorem AH.27); 1.42% leading-order mean error
with seven data-selected prefactors (main volume,
App. FM).

• α = 1/137.036: Chern-Simons + Toeplitz at
kmax = 60 (Theorem AH.26); the identification
of its hypercharge-trace coefficient with 1/e2 is not
derived (main volume, the convention-locked closed
form for α).
• Ngen = 3: a stated input, numerically χtop (CP 2 ) =
3 (main volume, App. F).
The theorem is the conjunction of these dimensionlesssector statements; it does not override the normalization
ledger of App. FK. The remaining open frontier is the
origin of the primordial perturbations and the horizon
problem, which enter through one stated starting condition (main volume, App. Q) and for which the theory
states no mechanism; per Theorem AH.136, DFD has no
native inflaton. The pre-BBN optical engine would inflate
the optical metric only with a negative-pressure source
and a reheating stage that the printed theory does not
contain, and then only below H ≃ 4 × 1010 GeV (main
volume, App. Q), and the χ relic rules out any appended
stage that would leave a detectable tensor signal, CMAH
as stated included (r ≲ 10−7 ; Theorem AH.40).
Corollary AH.141 (Anthropic Principle Displaced).
The standard alternative to “why these constants?” is
the anthropic principle plus multiverse selection. DFD’s
structural determination provides a deterministic mechanism: the hierarchy exponents are fixed by derived or
discretely selected structure, not selected from an ensemble.
Proof. By Theorem AH.140, the dimensionless hierarchy
exponents and ratios are fixed by derived or discretely
selected structure; the continuous inputs of App. FK
remain. Predictivity scoreboard: windows allow ranges
(e.g., α in [1/200, 1/100] for chemistry); the closed form
gives 1/137.036 specifically. Five orders sharper.
f.

Holographic Principle: Photon-Sphere Bulk-Boundary
Duality

Theorem AH.142 (Bulk-Boundary Duality at Any Compact Mass, conjectural). For any compact mass M in
DFD, the bulk theory (flat R3 × Rt with ψ̂, ĥTT
ij , microsector) is dual to a 2D SU (2)58 WZW boundary CFT on
the photon-sphere ∂Brph at the areal radius 3GM/c2 of
DFD’s exterior, with central charge cBD = 174/60 = 2.9.
Operator-operator bijection:
O∂ (σ) = lim (r − rph )−∆ Obulk (r, σ).
+
r→rph

Proof. The SU(2)58 MTC, whose shifted level equals the
microsector dimension (Theorem AH.75), provides the
boundary CFT structure: 59 primaries with conformal
weights ha = a(a + 1)/60, central charge c = 3 · 58/(58 +
2) = 174/60 = 2.9. The photon sphere ∂Brph admits a
natural 2D CFT structure from the optical-metric induced
metric. Bulk-boundary correspondence: bulk operators

41
at radius r project to boundary operators with conformal
weight ∆a via the optical-radial scaling. Reconstruction
kernel in HKLL-style:
Z
ϕbulk (r, σ) = dAopt (σ ′ ) K DFD (r, σ; σ ′ )ϕ∂ (σ ′ ),
with K DFD (r, σ; σ ′ ) = (n(r)/n(rph ))−∆a Pa (cos γ) where
γ is the angular separation. Valid outside the event horizon.

g.

CKM ρ̄: Both Paths Recorded

Remark AH.143 (Apex paths A and B: status open). Neither Path A (+5/2 Cayley correction ⇒ ρ̄ = 21.5α =
0.157) nor Path B (integer 22 reassignment ⇒ ρ̄ = 22α =
0.161) is data-independent. Both routes use ∼ 1 bit of
PDG-driven selection from a structurally-admissible set.
Path A has deeper structural grounding (4 convergent
perspectives: Z3 cohomology, MTC SU(2)58 Verlinde,
Connes Spinc canonical shift, Cayley graph (A5 , S) gen-1
distance) and preserves the uniform selection rule across
channels. Path B has cleaner arithmetic (0.91% vs 1.39%)
but breaks the bundle-filtration rule. Path A is kept as a
“Conjectural Selection-Rule Extension”, with Path B crossreferenced as the alternative. Discrimination is experimental: hyper-Belle / FCC-ee at σ(ρ̄) ≤ 0.0012 will decide at
> 3σ. The integers 31, 108 and 49 of λ, A and η̄ are kept
as suggestive of cohomological origin; the strengthened
branch of Appendix AO uses (31, 108, 43/2, 49)α, whose
apex is the Path A value.

h.

QECC: Explicit Pauli-Stabilizer Construction

Theorem AH.144 (Heisenberg-Weyl Stabilizer Subgroup of M60 (C)). The microsector algebra M60 (C) contains a Heisenberg–Weyl subgroup HW(60) = ⟨X60 , Z60 ⟩
generated by the Z60 clock and shift pair, with abelian
3
stabilizer subgroup SDFD = ⟨Z60
⟩ of order 20. The
joint +1 eigenspace equals ker DF = span{|0⟩, |20⟩, |40⟩},
the 3-generation logical subspace. Logical Pauli pair
20
(X̄, Z̄) = (X60
, Z60 ) acts as qutrit Pauli on the 3-dim
code.
Proof. The cyclic group Z60 provides clock and shift generators X60 , Z60 acting on C60 with X60 Z60 = ωZ60 X60
where ω = e2πi/60 . Order: both have order 60. Abelian
3
subgroup: Z60
has order 20 and commutes with itself trivially. Joint +1 eigenspace: vectors |n⟩ with ω 3n = 1,
i.e. n ∈ {0, 20, 40} — 3-dimensional. This matches
20
ker DF = C3 . Logical operators: X̄ = X60
permutes
{|0⟩, |20⟩, |40⟩} cyclically (qutrit shift); Z̄ = Z60 phases
them by ω 20n = e2πin/3 (qutrit clock). X̄ Z̄ = ω 20 Z̄ X̄ =
e2πi/3 Z̄ X̄ — qutrit Pauli algebra.
Remark AH.145 (Caveats). The construction is a
measurement-theoretic Gottesman stabilizer code; it is
not a Hamiltonian Kitaev code because [DF , SDFD ] ̸= 0

as operators (only subspace preservation holds). Code
distance d = 1 in the standard HW-displacement metric,
d ≥ 4 in the Toeplitz-mode-jump metric. The literal
[[60, 3, ≥ 4]] reading holds in the Toeplitz metric only.
Lemma AH.146 (Evaluation-reading dichotomy: every embedding yields a = 9 or fails). Let s1 , s2 , s3 span
ker DF and let ι : ker DF ,→ H 0 (CP 2 , O(9)⊕O⊕5 ) be any
embedding into the written microsector ambient. Write
evx (si ) = (pi (x), ci ) ∈ O(9)x ⊕ C5 . Then exactly one of
two horns holds: (i) rank{c1 , c2 , c3 } = 3, in which case
ev is fibrewise injective, the evaluation subbundle is trivial O⊕3 , its determinant twist vanishes, and the charged
determinant line is det(ker DF ⊗ LY ) = O(9) identically
— the same answer as the fixed-kernel reading; or (ii)
rank{ci } ≤ 2, in which case some nonzero combination
is a pure O(9) section, vanishing on a nonempty degree-9
curve, so constant rank fails and no rank-3 evaluation
subbundle exists (the all-in-O(9) assignment favored by
the determinant-line bookkeeping is the extreme rank ≤ 1
sub-case). Hence a = 9 in every well-defined reading.
The lone variant that escapes both horns — saturating the
rank-2 case to the subsheaf O⊕2 ⊕ O(9), giving a = 18,
kmax = 195 — abandons the constant-rank subbundle
definition and reproduces exactly the q1 = 6 alternative
already excluded by the measured α (App. F, Rem. on
independence).
Remark AH.147 (Basis convention). For definiteness we
adopt the lexicographic dictionary: indices 0–54 of the
microsector basis label the 55 degree-9 monomial sections
of O(9), indices 55–59 the five trivial matter-type lines —
consistent with the determinant-line bookkeeping (App. F:
the generations are carried inside O(9); the O⊕5 are
the matter-type lines). Under this convention ker DF =
span{|0⟩, |20⟩, |40⟩} lies in the O(9) column and horn (ii)
applies in its rank-≤ 1 form.

i.

Microsector Mass Spectrum: Direct-Detection Locked Out

Theorem AH.148 (Microsector Mass Lattice). The 60
microsector modes split as 3 chirality zero modes (the SM
generations) plus 57 nonzero Toeplitz–Dirac eigenvalues
forming a near-arithmetic lattice
m2n ≈ (n − 3) · Λ2top /57,

n = 4, . . . , 60.
√
The lightest nonzero mode is m4 ≈ Λtop / 57 ≈ 1.4 ×
1017 GeV, thirteen orders above the LHC and twelve above
a 100 TeV collider.
Proof. The Spinc Dirac operator DF on CP 2 × S 3 at
kmax = 60 has 3 zero modes (kernel) plus 57 positive eigenvalues√distributed approximately uniformly
be√
tween ∼ Λtop / 57 and Λtop . With Λtop = αMP =
17
1.04 × 1018 GeV: m4 = 1.4
p× 10 GeV. No mode is
sub-TeV. The bound m4 ≥ α/kmax · MP holds for the
printed spectrum, which the rigidity premise (layer 7 of
Theorem AH.179) holds fixed.

42
Corollary AH.149 (No Direct Lab Production). No microsector mode is producible at any current or planned collider. The best analog test is a 60-state condensed-matter
system on trapped-ion or superconducting hardware.
FCC
Proof. Direct: m4 ≫ Ecm
≈ 100 TeV.

j.

Amplitude Bootstrap Saturation

Theorem AH.150 (DFD Saturates Minimum-Spectral-Gap Bootstrap Inequality, conjectural). Among SMcompatible UV completions, DFD saturates the minimumspectral-gap p
inequality: the lightest non-zero state has
mass mgap =
α/57 MP , the smallest gap consistent
with Spinc integrality and (3, 2, 1) gauge partition. The
2
leading amplitude EFT coefficient c3 on the Rµνρσ
operator is c3 = 1/6 forced by two independent derivations:
(i) Witten degree-1 twistor-string symmetry factor 1/3!.
(ii) One-loop Toeplitz quadrature on Hmicro (Theorem AH.37).
√
√
Proof. Spectral gap: mgap = m4 = αMP / 57 from
Theorem AH.148; this is the minimum gap consistent
with dim Hmicro = 60 and the Spinc index. Each of (i)
and (ii) gives c3 = 1/6 via different combinatorics. The
convergence is non-trivial because the two derivations use
distinct categorical structures.
Corollary AH.151 (Primordial gravitational-wave helicity asymmetry). The c3 = 1/6 coefficient gives a primordial gravitational-wave background the helicity asymmetry
Ah = c3 (H/MP )2 at the rate H of the stage that froze it.
The χ relic caps that rate at H ≲ 4 × 1010 GeV (main
volume, App. Q), where Ah ≲ 2 × 10−18 , and the background itself is r ≲ 10−7 (main volume, App. SM); no
helicity signal is predicted for LISA.
Proof. At H = 4.3 × 1010 GeV, (H/MP )2 = 1.2 × 10−17
and Ah = 2.1 × 10−18 . A rate H ∼ 10−5 MP would give
Ah = 16 (10−5 )2 = 1.7 × 10−11 , and the relic excludes it
(computed).

k.

Dark-Matter Null Portfolio

Theorem AH.152 (DFD Predicts Null Results Across
45 DM Searches). DFD predicts zero signal across
all current-reach dark-matter direct-detection, indirectdetection, axion-haloscope, sterile-neutrino, hiddenphoton, and modulation experiments. DFD’s cold dark
matter is the derived χ-matter field (App. AV), a nonthermal, Standard-Model-neutral pseudoscalar with no
photon coupling (its photon coefficient is E = 0, derived
in App. AV); it sources none of their recoil or haloscope
channels. The set of null predictions therefore covers every
current search (∼ 45 experiments). With mχ ≃ 32 eV and
E = 0, χ has no decay line and no conversion line: any

dark-matter photon line at 16.0 eV (77.5 nm, two-photon
decay) or axion-like signal at 32.005 eV (38.74 nm, conversion), at any strength, falsifies it. This breadth makes
DFD multiply falsifiable: a confirmed halo-recoil signal in
any current-reach channel falsifies it.
Proof. DFD has no postulated BSM particle dark matter
(Theorem AH.64, items i–vi); its cold dark matter is the
derived χ-matter field (App. AV), the harmonic b3 threeform on S 3 = SU (2), which is Standard-Model-neutral
and non-thermal and so produces no recoil or haloscope
signal in the channels below. Galactic phenomenology
√
is separately covered by the µ-law with a0 = 2 α cH0loc
(Theorem AH.21; main volume, App. ZS); inside galaxies
the two compose by the adopted floor law with a stated
three-dimensional filling rule (App. AV). The chiral index dim Hmicro = 60 is saturated by SM matter; χ is
a bosonic harmonic mode, not one of the 60. DFD’s
σ8 of record, 0.806, is that of the cold χ carrier with
Qχ = 1 (main volume, App. GR); the value 0.811 from
Ωb with the optical-metric amplification QDFD ≈ 502
(Theorem AH.7) belongs to the baryon-only construction
of the main volume’s App. AE, which is not DFD’s growth
carrier. Null prediction is structural across: xenon TPCs
(XENON-nT, LZ, PandaX, DARWIN), argon (DEAP3600, DarkSide), spin-dependent (PICO-500), low-mass
(SuperCDMS, DAMIC-M, SENSEI), modulation (DAMA,
ANAIS, COSINE, SABRE-South), axion (ADMX, MADMAX, IAXO), sterile-ν (KATRIN, JUNO, DUNE, MicroBooNE), hidden-photon, indirect γ-ray (Fermi-LAT,
HESS, MAGIC, CTA), cosmic-ray (AMS-02, GAPS), neutrino (IceCube-Gen2, KM3NeT).
Corollary AH.153 (Single-Statement Falsifier). A single confirmed positive detection of any halo-DM signal
(cross-experiment confirmed, > 5σ, halo-consistent recoil/coupling) at any of the 45+ current-reach experiments
falsifies DFD’s current-reach-null closure at theorem level
(the derived χ-matter field, with no photon coupling, predicts no such signal).
Proof. DFD’s claim is single-statement: zero halorecoil/haloscope signal from any thermal WIMP or
current-reach axion-like particle. The derived χ-matter
field has ρχ ̸= 0 (Ωχ h2 ≃ 0.12; App. AV) but is nonthermal and Standard-Model-neutral, so it sources none of
these channels; its cold clustering is tested cosmologically
(CMB third-peak height), not in direct detection. Any
cross-confirmed such recoil detection contradicts DFD.
ΛCDM by contrast can absorb any null result by shifting
search parameters; DFD cannot.
l.

Axiomatic Foundation

The inputs of the graded predictions, the postulates,
adopted axioms, identifications, discrete selections and
continuous calibrations, are listed in the input tables of
App. FK of the main volume.

43
10.

Additional Theorems and Proofs (Series V)

This subsection covers the master-action restatement,
the single-invariant theorem and the two-axiom minimum
lemma.

a.

Master DFD Lagrangian: Single-Equation Statement

Theorem AH.154 (Master DFD Action). The microsector, gauge and Einstein–Hilbert content of DFD is
organized by the Connes-style spectral action
SDFD = Tr f (D/Λ⋆ ) + ⟨Ψ, DΨ⟩
where D = DM4 ⊗ 1 + γ5 ⊗
√DF is the full Dirac operator
on M4 × CP 2 × S 3 , Λ⋆ = α MP , f is a heat-kernel cutoff, and dim HF = 60. The expression fixes the claimed
dimensionless microsector structure with no continuous
shape fit; conversion to the physical cosmological prediction uses the four cosmological calibrations of App. FK of
the main volume (the vacuum coefficient R, the amplitude
As , the ωb coefficient, and τ ).
Heat-kernel
expansion
Proof.
P
2
4−2n
a
(D
)f
Λ
gives:
n
n n

Tr f (D/Λ)

=

• a0 coefficient: a Planckian vacuum term (App. WS);
reading it as the α57 term of Theorem AH.23 is
a conjecture that needs the whole DF -dependent
vacuum suppressed with it.
• a2 coefficient: Einstein-Hilbert with G
as the Newton constant.

−1

∝ αMP2

• a4 coefficient: Yang-Mills terms for the gauge
fields of the inner fluctuations, whose group

SU (3) × SU (2) × U (1) is selected, not forced (Theorem AH.214), + Higgs potential (Theorem AH.17).
The conjectured λH = 1/8 of the main volume
(App. Z) is stated at an electroweak matching scale,
not at Λ⋆ , where the measured Higgs mass needs a
negative quartic, −0.011 ± 0.003 (computed).
• ⟨Ψ, DΨ⟩: Dirac kinetic + Yukawa from off-diagonal
DF via Berry-bundle overlap integrals (Theorem AH.27).
The 3 generations are a stated input identified with
χtop (CP 2 ) = 3 (main volume, App. F). θ̄QCD = 0 (tree
level; loop-induced ≲ 10−16 ) from the det-orthogonal treelevel protection together with η-invariant vanishing on the
CP mapping torus, with θbare = 0 fixed by the spectral
action given that nothing is added to it for the gauge
fields (Theorem AH.66). The proof aggregates all prior
theorems.

b.

Single Master Topological Invariant

Theorem AH.155 (One Integer Organizing the Hierarchy Exponents). The printed hierarchy exponents of
the Standard Model and the cosmological constant are
organized by a single topological integer
Ω ≡ dimC Hmicro = χ(CP 2 , O(9) ⊕ O⊕5 ) = 60,
together with one external dimensional anchor (e.g., MP
or H0loc ), with the inputs listed in App. FK of the main
volume.

Proof. The integer Ω = 60 arises in four ways: (a) Hirzebruch-Riemann-Roch index on CP 2 with bundle E; (b)
icosahedral group order |A5 | = 60; (c) E8 root system |RE8 |/4 = 240/4 = 60; (d) shifted level k + h∨ = 58 + 2 of the
SU (2)58 MTC. With the inputs of App. FK of the main volume, and with θbare = 0 fixed by the spectral action given
that nothing is added to it for the gauge fields, the printed formulas read:
α−1 = (π 3/2 /24) · 10 · Ω · ((Ω + 3)/(Ω + 4)) · [1 + 7/(80 · 4095)] = 137.036,

7/(80 · 4095) = 1/(13 Ω2 )

GℏH02 /c5 = CαΩ−Ngen = Cα57
√
v/MP = α8 2π mR /MP = α3
√
mf = Af αnf v/ 2 (Cayley graph on A5 )
θ̄ = 0

c.

(tree level; loop-induced ≲ 10−16 )

Wheeler-DeWitt Equation in DFD

Theorem AH.156 (Wheeler-DeWitt Closed-Form
Ground State). The DFD Wheeler-DeWitt equation
ĤDFD Ψ = 0 (equivalently Ĥ ′ Ψ = E0 Ψ with E0 =

Ngen = 3 (input; numerically χtop (CP 2 )) .

α57 MP4 under the additive-constant convention of Theorem AH.177 — under the lock of the lapse to the scale
factor a constant added to the Hamiltonian acts as a curvature term (w = −1/3), not as vacuum energy, so the
two notations agree only when E0 is carried as a vacuum

44
term of the action (computed); that theorem’s eigenfunction and eigenvalue steps are asserted, not derived, per
its status note, §AH 12 b) has the proposed ground state
Ψ0 [ψ, Ξ] = N exp(−SE [ψ, Ξ]/ℏ),
the Boltzmann weight of the Euclidean DFD action. The
exponent of the cosmological constant appears in the
Wheeler-DeWitt eigenvalue α57 MP4 , about 14 times the
observed density (scope note of Theorem AH.177); the
ground state fixes no amplitude for the primordial perturbations (main volume, App. FK).

Proof. DFD’s flat spatial arena gij = δij is non-dynamical,
and in the homogeneous minisuperspace the traceless
strain of the parent tensor vanishes by symmetry; only
ψ̂ and the 60-dim microsector are quantized. This eliminates the supermetric pathologies of standard WheelerDeWitt. The constraint reduces to an elliptic functional eigenvalue problem on L2 (µβ ) with respect to the
Parisi-Wu invariant measure. Strong monotonicity of
µ(x) = x/(1 + x) (App. U) provides coercivity; the status note of Theorem AH.177 records the gaps. Direct
check: ĤDFD N e−SE /ℏ = 0 via the standard Schrodingeraction Euclidean correspondence. Cosmological constant
emerges as the lowest eigenvalue of the FRW-restricted
Hamiltonian, computed via App. O determinant ratio.
Vacuum ψ fluctuations on the scale of today’s horizon are
of order H0 /MP ≈ 1.2 × 10−61 , far below the observed
10−5 , which has no mechanism in DFD (main volume,
App. FK).

d.

Single-Vacuum Theorem (No Landscape)

Theorem AH.157 (DFD Has Exactly One Vacuum).
The DFD vacuum manifold has dimension zero modulo
gauge transformations and discrete symmetries:
dim(VDFD / ∼) = 0.
There is no DFD landscape.
Proof. Five-step composition of prior closure theorems:
(1) manifold CP 2 × S 3 is unique (Theorem AH.9); (2)
bundle E = O(9)⊕O⊕5 follows from the determinant-line
arithmetic at the derived q1 = 3 (main volume, App. F,
the remark on the flux quantum); (3) the Toeplitz cap
kmax = 60 follows from Ngen = 3, q1 = 3 and the five
hypercharged multiplet types (main volume, App. F); (4)
the ψ vacuum is unique up to the constant shift of its flat
direction (the scalar action carries no potential term in ψ;
Theorem AH.136, sub-(i)); (5) microsector ground state
is unique by App. F gauge emergence, with the internal
Dirac operator held at its printed spectrum by the rigidity premise (layer 7 of Theorem AH.179); without that
premise the inner fluctuations of M60 (C) would make its
60 eigenvalues dynamical (main volume, App. FK). Ten
classes of counter-vacua (alternative manifolds, alternative bundles, alternative complex structures, alternative
Toeplitz truncations, alternative gauge partitions, alternative Spinc structures, alternative metric deformations,
alternative boundary conditions, alternative microsector
fillings, alternative ψ vacua) are excluded given the stated
selections and the rigidity premise.
Comparison with string theory: dim(Vstring ) ∼
500–1000 continuous moduli plus ∼ 10500 discrete choices;
DFD: 0 continuous + 1 discrete. The string-landscape
problem does not arise in DFD.

e.

Vacuum Deposit: All-Orders Theorem

Theorem AH.158 (All-Orders Bound on the Vacuum Deposit).



3 57 4 
α2
ρdep =
α MP 1 + R(α) ,
|R(α)| ≤
+ exp −8π 2 /α ≈ 5 × 10−5 .
2
8π
1−α
This is the C = 1 deposit; the observed dark-energy density is (0.6042 ± 0.0027) ρdep , its coefficient measured and
underived (main volume, App. AU). The 10122 fine-tuning problem is closed at theorem grade in its exponent, for
the microsector vacuum piece. (Scope: the bound governs the 60-mode microsector trace only; the Standard-Model
matter-sector vacuum loop — external to the microsector, of order α32 MP4 — is not bounded by this theorem, nor is
the Planckian a0 vacuum term of the spectral action (Theorem AH.154), and no full-spectrum vacuum sum rule is
derived. The matter loop and the a0 term are each assumed to cancel, and the matter-piece cancellation remains the
universal open cosmological-constant problem; see the scope remark in App. AU.)
(0)

Proof. Tree level: ρdep = (3/8π)α57 MP4 from Theorem AH.23. One-loop: |δΛ(1) /Λ| ≤ α2 from Wilsonian regulation
(Theorem AH.1). n-loop bound: |δΛ(n) /Λ| ≤ α2n by induction with per-loop suppression from finite microsector
trace. Geometric series sums to α2 /(1 − α2 ) ≈ 5 × 10−5 . Non-perturbative contributions: instantons of action
Sinst = 8π 2 /α ≈ 10820 give e−10820 ≈ 10−4699 , negligible. Domain walls forbidden by π2 (V) = 0. Sphalerons thermally
suppressed at T = 0. Total all-orders correction bounded by 5 × 10−5 .

45
f.

EW-Planck Hierarchy: No-Tuning Theorem

Theorem AH.159 (Hierarchy Problem Eliminated, Not
Solved). DFD does not solve the hierarchy problem by
tuning a parameter; it eliminates the bare-parameter form
of the question by deriving v from topology: the bare
parameter m2H does not exist in DFD’s parameter list.
Radiative corrections to v from the microsector satisfy
|δv/v| ≤ O(α2 ) ≈ 5 × 10−5 at all loop orders. (Scope:
the external 4D Coleman–Weinberg loop over noncompact
momenta is not regulated by the microsector trace — see
the status note of §AH 2 a — so radiative stability of the
derived v against the external matter loop remains the
universal open hierarchy problem, shared with all QFT
coupled to gravity; what is eliminated is the tuned-bareparameter formulation.)
√
Proof. v = MP α8 2π: the√exponent 8 is forced (dim X +
1, Theorem AH.17); the 2π prefactor is fitted, and
no printed rule fixes it (main volume, App. K). The
Higgs mass is m2H = 2λH v 2 , with λH = 1/8 a dimensioncounting conjecture (main
volume, App. Z). Wilsonian
√
regulation at Λtop = αMP (Theorem AH.1) plus permode α-cancellation (Lemma AH.2) gives |δv/v| ≤ O(α2 )
at one loop. All-orders sum: convergent geometric series
α2 /(1 − α2 ). Compare with SUSY: SUSY cancels δm2H
via symmetry, requires TeV superpartners (LHC null),
suffers a ”little hierarchy” 10−2 –10−4 . DFD eliminates the
parameter m2H itself; no new physics below Λtop ≈ 1018
GeV.

g.

All-Orders QG Amplitude Finiteness

Theorem AH.160 (Universal Amplitude Finiteness, conjectural). Every DFD scattering amplitude at every loop
order L with every external leg count n is finite.
Proof. Six-step proof: (1) Hmicro is 60-dimensional
via Atiyah-Singer; (2) continuum 4-momentum integration is bounded by Λtop (no Landau pole, Theorem AH.16); (3) combined integration is a finite sum
of finite terms (Toeplitz quadrature); (4) per-loop bound
is sub-exponential; (5) UV divergences absent (no GoroffSagnotti R3 counterterm because integrand is finite trace,
not divergent loop integral); (6) IR divergences cancel by
standard Bloch-Nordsieck/KLN. RG flow freezes above
Λtop .
Corollary AH.161 (DFD vs String Theory UV-Finiteness). DFD and string theory approach UV-finiteness
by different mechanisms: string theory uses an infinite
Regge tower with modular invariance; DFD uses a finite
tower of 60 microsector states. DFD’s gravity sector carries one measured coefficient, the early-clock coefficient
C = 1/λ2 (main volume, Apps. FK and ZS); string has
compactification moduli at ∼ 500–1000.
Proof. Parameter counting: see Theorem AH.157.

h.

Quantum Measurement Problem, Structural Reduction

Theorem AH.162 (Structural Reduction of the Measurement Problem). In DFD, the optical metric n = eψ
is sourced by the c-number expectation ⟨Ψ|ρ̂|Ψ⟩ via an
elliptic PDE with monotone µ, giving a unique optical
metric. No metric branching occurs; the Many-Worlds interpretation is structurally excluded. The Born-rule form
follows from Gleason’s theorem once the complex structure
J⋆ is supplied, and its value is conditional on J⋆ (Theorem AH.178); the measurement postulate is imported
(main volume, App. FK).
Proof. By Theorem AH.24, the elliptic equation has a
unique weak solution ψ for any L2loc source (Browder–
Minty with strict monotonicity of µ(x) x; main volume,
App. U). Hence the optical metric is unique. There is no
branching. Decoherence in the position basis follows from
the standard ρ-coupling in the optical-metric prescription;
pointer states are position eigenstates. Born rule: Gleason’s theorem fixes the form of the measure on the projection lattice of Hmicro (dim = 60 ≥ 3) once J⋆ supplies
the inner product; the value is conditional on J⋆ (Theorem AH.178). Tree-level intrinsic gravitational collapse is
τ = ∞ (opposite of Diosi-Penrose τ ∼ ℏ/Egrav ).
Falsifier: MAQRO experiment (10−13 kg test mass,
space-vacuum, 100 s coherence time): DFD predicts sustained interference fringes; Diosi-Penrose predicts collapse
at 10−3 s. This sustained-fringe prediction is the direct
consequence of the no-branching theorem above (one cnumber-sourced optical metric, no per-branch self-force),
consistent with the open-problems section; the per-branch
deep-MOND-enhanced collapse alternative is excluded by
the two-body structure of the derived vertex (main volume, App. ZU).

i.

Direct Microsector Lab Analog

Theorem AH.163 (Quantum-Hardware-Buildable DFD
Microsector Simulator). A [[60, 3, ≥ 4]] stabilizer code
(distance ≥ 4 in the Toeplitz metric, 1 in the standard
displacement metric; Theorem AH.181) implementing
DFD’s microsector structure can be constructed on existing quantum hardware (Quantinuum H2-1 trapped ions,
IBM Heron, or Google Willow).
Proof. Map each microsector mode to a qudit basis state
in a 60-dim Hilbert space. Implement Spinc Dirac
DF as a block-diagonal Hamiltonian with eigenvalues
{0, 0, 0, λ4 , . . . , λ60 } via a 19-laser AC-Stark protocol. Sta3
bilizer measurement of ⟨Z60
⟩ via Knill ancilla phasekickback.

46
j.

AdS/CFT Without AdS: A Photon-Sphere Holographic
Duality

Theorem AH.164 (Photon-Sphere Holographic Duality
on Flat Spacetime, conjectural). For any compact mass
M in DFD with photon sphere at the areal radius 3GM/c2 ,
the bulk partition function with Dirichlet data ∂ϕ on ∂Brph
equals the SU (2)58 WZW boundary partition function with
source J = ∂ϕ:
SU (2)58

∂
Zbulk [gij
, ∂ϕ] = Z∂

[J(σ) = ∂ϕ(σ)],

invariant under D → λD with Λ → λΛ (rescaling). Hence
the dimensional scale cannot be fixed by any structural
axiom alone; one anchoring axiom is needed.
Two-axiom minimal statement:
(SA) Structural: DFD = spectral action of the real
KO-dim-4 spectral data (Theorem AH.77) over
C ∞ (R3 ×Rt )⊗End(Berezin-Toeplitz Hilbert space)
for a compact Spinc Kähler-times-Lie-group manifold with χtop (Mc ) = 3, positive Ricci curvature,
π1 (Mg ) = 1, π3 (Mg ) = Z and minimal dimension.

+
to all orders in the asymptotic expansion as r → rph
. The
boundary central charge is cBD = 174/60 = 29/10 = 2.9.

(AA) Anchoring:
One
{MP , v, H0loc , . . .}.

Proof. The induced optical metric on ∂Brph carries an
emergent SO(3, 1) Möbius conformal structure intrinsic
to the round S 2 with Weyl factor e2ψ(rph ) = 9 on the static
exterior, where eψ = (1 − 2GM/c2 R)−1 = 3. This conformal symmetry is not inherited from any bulk isometry;
it lives on the boundary. The boundary CFT is SU (2)58
WZW (level k = 58, central charge c = 3k/(k + 2) = 2.9).
Bulk fields project to boundary primary operators of
conformal weight ∆a = a(a + 1)/60 via radial scaling; HKLL-style reconstruction kernel K DFD (r, σ; σ ′ ) =
(n(r)/n(rph ))−∆a Pa (cos γ) valid outside the event horizon.
Partition-function identity follows from Witten-style bulkboundary correspondence applied to the optical-metric
path integral.

Comparison with Einstein-Hilbert: GR’s effective
axiomatic rank is also 2 (4-diff-invariant scalar action +
Newton’s G).

Corollary AH.165 (Local vs Global Holography). DFD
has local holography (one CFT per object); AdS/CFT
has global holography (one boundary CFT for the whole
bulk). DFD’s boundary CFT is strictly more rigid: k = 58
fixed by χ(CP 2 , E) = 60, no free parameters versus AdS’s
continuous N, R, gs .
Proof. Direct: each photon sphere supports its own boundary CFT independently; AdS supports a single global one.
Rigidity comparison: count parameters.
k.

Two-Axiom Minimal Statement of DFD

Theorem AH.166 (Single-Axiom Reduction Impossible; Two-Axiom Minimum). The DFD theory cannot be
reduced to a single axiom (conformal self-similarity of
spectral triples blocks it: D → λD with Λ → λΛ gives
an isomorphic spectral action). The minimum axiomatic
rank for the internal spectral object is two: one structural
axiom selecting the spectral object, one anchoring axiom
selecting the dimensional scale. (The complete theory
additionally posits the flat R3 arena and the optical map
n = eψ , and imports the quantum seed — the numeric ℏ
and an indifference/typicality bridge for Born; the global
irreducible-input count is therefore larger. The “two” is
the irreducible rank of the internal-index sublayer, not of
the whole framework.)
Proof. Single-axiom impossibility: any structural axiom characterizing the DFD spectral data (A, H, D) is

11.

external

scale

Λ⋆

∈

Additional Theorems and Proofs (Series VI)

This subsection treats precision-frontier overdetermination and comparisons with rival quantum-gravity programs.
It contains: (i) a closed-form α−1 at Ω = 60, exact under
two stated choices, whose integers alone lie 21 ppm from
CODATA 2022 and which lies 0.005 ppm from it with the
open bracket coefficient; (ii) sub-meV neutrino spectrum
closure with NuFIT 6.0 χ2 (2) = 0.021 at zero continuous
parameters (one data-selected branch choice among four
equally natural fractions; see the branch-selection remark
of Appendix X in the main paper); (iii) a +0.23% (α/π)
sector-resolved Casimir coefficient, a conjecture outside
DFD’s claims (Theorem AH.171); (iv) a comparative documentation block listing concrete empirical falsifiers for
DFD that the four rival programs do not currently predict at theorem grade, with the scorecard presented as
documentation, not as a uniqueness or superiority claim;
(v) a cross-check of 27 live dual-route pairs, each agreeing
below 1σ.
a.

Closed Form for the Fine-Structure Constant

Theorem AH.167 (Closed-form α). The closed-form
polynomial in the master invariant Ω,
i
π 3/2
Ω+3h
7
α−1 (Ω) =
Tr(Y 2 ) Ω
1+
,
2
24
Ω+4
80(Ω + 4) − 80
evaluated at Ω = 60 with Tr(Y 2 ) = 10 on the (3, 2, 1)
−1
gauge-bundle structure of CP 2 × S 3 , gives αDFD
=
137.03599985412 . . .. The CODATA 2022 measured
value is α−1 = 137.035999177(21). The residual is
−1
−1
αCODATA
− αDFD
= −6.8 × 10−7 (−0.0049 ppm, matching the convention-locked −0.005 ppm statement of the
main paper); this is ∼ 32× CODATA’s quoted uncertainty of 2.1 × 10−8 , i.e. not an agreement within current
experimental error.

47
Proof. The π 3/2 /24 prefactor is the heat-kernel normalization of the main volume’s exact construction: the
ζ(0) of the CP 2 charge trace, −1/24 per √
unit charge
squared, times the terminating heat trace ( π/4)Λ3 of
the ρ-shifted Casimir on S 3 , with the heat-kernel regulator. It is not a Spinc Hirzebruch–Riemann–Roch number,
since those are rational. The Tr(Y 2 ) = 10 factor counts
the standard-model hypercharge invariant on the threegeneration (3, 2, 1) bundle. With the standard YQ = 1/6
normalization (Slansky 1981 p. 96), Tr(Y 2 )|one gen =
6 · (1/6)2 + 3 · (2/3)2 + 3 · (−1/3)2 + 2 · (−1/2)2 + 1 · (−1)2 =
1/6 + 4/3 + 1/3 + 1/2 + 1 = 10/3; summed over three
generations gives 10; its identification with 1/e2 rather
than with the hypercharge coupling is not derived (main
volume, the convention-locked closed form for α). In the
factor Ω(Ω + 3)/(Ω + 4), the ratio (Ω + 3)/Ω is the exact
quantum-to-classical ratio of the CP 2 hypercharge sec-

ond moment and Ω + 4 is the Toeplitz dimension (main
volume, convention-locked closed form). The coefficient
of the 7/[80(Ω + 4)2 − 80] correction is a stated rational,
1/13 at Ω = 60, whose provenance is open: the CP 2
Toeplitz sector is exact at every level and supplies no such
term. Substituting Ω = 60, Tr(Y 2 ) = 10, Ω + 4 = 64,
80 · 642 − 80 = 327600 yields the displayed value to 14
digits via standard arbitrary-precision arithmetic.
The closed form has no continuous parameter, but
its last 21 ppm rests on the bracket coefficient, whose
provenance is open, and the 5 ppb figure is a property of
the closed form rather than a test of the bracket.

b.

Sub-meV Neutrino Spectrum Closure

Theorem AH.168 (Sub-meV neutrino-spectrum closure). Branch-B exponents k = α−3/11 , r = α−7/20 together with
the 14/13 finite-dimensional priming factor on the right-handed Majorana scale MR = MP α3 predict the absolute
neutrino mass spectrum at sub-meV precision:
m1 = 2.342 meV,
Σmν = 61.462 meV,

m2 = 8.963 meV,

mβ = 9.1514 meV,

m3 = 50.157 meV,

mββ ∈ {0.13, 2.33, 3.25, 5.45} meV,

with normal ordering, and δCP = −π/2 from the µ–τ reflection, a structural postulate (main volume, App. FK); the
four mββ values are the discrete Majorana-phase branches of the canonical TM1 chain. Confrontation with NuFIT
6.0 gives pull21 = −0.031σ, pull31 = −0.141σ, χ2 (2) = 0.021, p = 0.99. DFD lands precisely on the global-fit best
point with zero continuous parameters. (The rounded-prediction evaluation of Appendix X gives χ2 = 0.025; 0.021 is
the full-precision-spectrum value — both at zero continuous parameters. Branch-selection status: the A–B binary is
data-selected per the branch-selection remark of Appendix X of the main paper.)
Proof. The seesaw composition mν ∼ m2D /MR with Dirac-Yukawa mD = MP αk and Majorana scale MR = MP αr
yields mν = MP α2k−r at tree level. Branch B fixes k = α−3/11 exponent and r = 3 via Theorem AH.28. The 14/13
priming factor is dν /(dν − 1) with dν = a + 5 = 14 (main volume, App. X). Numerically: m3 = (14/13) π MP α14 =
(14/13)·(1.221×1019 GeV)·(3.815×10−30 ) = 50.157 meV, with the mass ratios m2 /m1 = α−3/11 and m3 /m2 = α−7/20 .
The lower masses follow from the diagonal entries of the PMNS matrix with θ12 = 34.35◦ , θ23 = 45.0◦ (maximal; the
µ-τ reflection forces θ23 = π/4 bundled with δCP = −π/2, a sharp falsifiable prediction in soft ∼2σ tension with NuFIT
6.0, octant unresolved), θ13 = 8.509◦ . The normal-ordering forcing follows because m3 > m2 > m1 in the geometric
sector decomposition (App. X.8) is incompatible with inverted ordering. The NuFIT 6.0 χ2 residual evaluates to 0.021
on two degrees of freedom (the two mass-splittings ∆m221 and |∆m231 |), giving p = 0.99.

Falsifier: JUNO’s mass-ordering determination at 3σ
(∼2032). Inverted ordering falsifies the entire microsector
architecture with no patch. A Simons Observatory +
DESI + Euclid measurement of Σmν > 80 meV at 3σ
(∼2032) likewise falsifies.
c.

Comparative Structural Profile of DFD vs. Four Major
UV-Completion Programs

This subsection documents DFD’s comparative structural profile against four major UV-completion programs
(string theory, loop quantum gravity, causal set theory,
asymptotic safety) on a fixed list of structural and empir-

ical criteria. It is presented as a comparative documentation, not as a uniqueness theorem; a uniqueness claim
would require an external precise definition of “candidate
unified theory.”
Proposition AH.169 (Comparative structural profile:
DFD vs. four UV-completion programs). The four leading
UV-completion programs (string/M-theory, loop quantum
gravity, causal set theory, asymptotic safety) are scored
against DFD on seven structural-and-empirical criteria
below. DFD scores positively on all seven; no listed rival scores positively on all seven. This documents the
comparative status; it is not a uniqueness theorem.
1. No continuous parameter fitted to a predicted shape

48
except the baryon density (main volume, App. FK).
2. No landscape problem (Single-Vacuum Theorem AH.157).

a proper subset of them. The criteria are author-selected
and not intended as a uniqueness claim. The comparative profile, not the uniqueness claim, is the substantive
content of this subsection.

3. A single spectral action for the microsector, gauge
and Einstein–Hilbert content (Theorem AH.154).
4. A single topological invariant (Ω = 60, Theorem AH.155).
5. A two-axiom minimal statement (Theorem AH.166).
6. The SM + Λ + H0 exponent ladder derived, with
the inputs of App. FK.
7. 25+ sharp empirical falsifiers within 25 years.
Proof. The seven criteria are individually established by
the cited theorems for DFD. The comparison is by direct
enumeration over the four rival programs.
vs. String Theory. String theory has ∼ 10500 vacua
(criterion 2: no); does not predict the SM gauge group
(3, 2, 1) structurally (subsumed under criterion 6: no);
does not predict fermion masses; does not predict α, Λ, or
H0 ; relies on selection. Has neither a single master action
that determines all SM constants (criterion 3: partial
only), nor a single topological invariant (criterion 4: no).
Its Yukawas depend on continuous moduli (criterion 1:
no). Number of theorem-grade falsifiers within 25 years:
zero (criterion 7: no). DFD scores positively on all 7
criteria where string theory scores positively on at most 1
(the master action partial credit).
vs. Loop Quantum Gravity. LQG does not structurally predict the SM gauge group, fermion masses, Higgs
mechanism, Λ, or α (criterion 6: no). No single master
action that derives the SM (criterion 3: no). LQG’s
Wheeler-DeWitt formal results and BH entropy from
spectral count are real, but the comparison on criterion
6 is decisive: LQG has not closed the SM continuousparameter prediction problem.
vs. Causal Set Theory. CST matches DFD on discrete cardinality at the fundamental level (kmax = 60
Toeplitz quantization vs. causal-set count)
and on a small√
number Λ outcome (Sorkin’s Λ ∼ 1/ N vs. DFD’s α57 ).
CST is the closest competitor on Λ alone but does not
extend to the full SM spectrum (no fermion masses, gauge
group, Higgs, or α at theorem grade); criterion 6: partial
only.
vs. Asymptotic Safety. AS shares DFD’s UVfiniteness ambition but assumes the existence of a nonGaussian fixed point (NGFP). AS has not delivered SM
closure (no fermion masses, Higgs, gauge group at theorem
grade).
A summary scorecard per criterion is straightforward;
positive scores are recorded only where the listed program
achieves the structural outcome at theorem grade. The
comparative profile (rather than a numerical scorecard) is
presented in qualitative form: DFD predicts each of the
seven listed criteria; the four rival programs each predict

d.

Cross-Cutting Consistency Theorem

Theorem AH.170 (Cross-cutting consistency of dual-route predictions). The historical register contains 28
cases in which two predictions reached the same observable
by different routes. Twenty-seven remain live and agree
at sub-1σ. The 28th was the optical screen route to H0 ,
which fails (see the proof ), so it is not a dual-route confirmation. Scope. This is a statement about the enumerated
pairs only. It does not assert that the corpus contains no
internal inconsistency of any other kind.
Proof. By the master-action theorem (Theorem AH.154)
and the single-invariant theorem (Theorem AH.155), the
live pairs descend from the common structural source
LDFD and Ω = 60. Explicit cross-check gives residuals
below 1σ for the 27 surviving pairs. The 28th comparison
had a 2.4σ residual, but its optical-screen H0 arm fails:
exact conformal invariance makes the global screen null,
and its C = 1 clock is the bound-matter clock of the twoclock postulate (main volume, App. ZS), not the early
clock. It remains in the historical enumeration only, not
in the live consistency score.
No single-number probability claim is attached to the
dual-route consistency check; predictions descending from
a common heat-kernel expansion are not statistically independent, so a joint-χ2 argument over them is not appropriate. The 27 live dual-route pairs are presented
as internal-consistency cross-checks rather than as an
aggregate-probability statement.

e.

Sector-Resolved Casimir Modification (Leading Order)

Theorem AH.171 (Leading-order sector-resolved
Casimir slope). DFD’s optical-metric coupling (n =
eψ ) implies a leading-order modification to the standard
QED Casimir force in a sector-resolved (metal/dielectric)
parallel-plate geometry of magnitude
∆F/F = α/π ≈ 2.32 × 10−3 = 0.232%,
arising directly from the sector-resolved optical-metric
coupling: gauge-field zero-point fluctuations couple via
the optical metric g̃µν = e2ψ gµν while material-bound
fluctuations couple via the matter metric ηµν .
Proof. The standard QED Casimir energy density between parallel plates at separation d is uQED =
−π 2 ℏc/(720d4 ). In DFD, the gauge-field vacuum responds
to g̃µν while the matter sector (responsible for the boundary conditions) responds to ηµν . For a sector-resolved
geometry at d ∼ 100 nm with local boundary-induced

49
refractive contrast ∆ψbdy ∼ α/(4π) (this is the standard
one-loop QED self-energy scale at the boundary), the
relative sector-resolved correction is
∆F/F = 4 ∆ψbdy = 4α/(4π) = α/π ≈ 0.232%.
This is the parallel-plate leading-order estimate.
The present prediction is the leading-order α/π =
0.232% value; the falsifier window is widened correspondingly to reflect the geometric uncertainty in the boundarycondition sector. A future explicit Lifshitz-formula derivation may sharpen the central value.
Status: The geometric prefactor of the boundary sector is unresolved, so the coefficient is a conjecture outside
DFD’s claims. A sub-0.1% precision Casimir-force measurement in sector-resolved (metal/dielectric) geometry
that finds a modification outside [+0.15%, +0.50%] refutes the conjecture, not the theory.

Proposition AH.174 (Moonshine connection, conjectural). The McKay–Conway–Norton ADE classification
chain |A5 | = 60 → E8 → Leech lattice → Monster situates Ω = 60 at the foot of the moonshine correspondence. Whether this is a deep mathematical fact (DFD
structurally encoding the moonshine–monster categorical
equivalence) or a numerological coincidence is open.
Proposition AH.175 (Atiyah–LeBrun–Witten triality, conjectural). DFD sits at the intersection of three
deep mathematical-physics structures: (i) the spectral side
(Connes spectral triple, Theorem AH.154); (ii) the twistor
side (the flag-manifold twistor space SU(3)/T 2 of CP 2 );
(iii) the categorical side (Reshetikhin–Turaev SU (2)58
TQFT on S 3 ). A rigorous triality theorem situating DFD
within Atiyah’s 2018 program, LeBrun’s twistor work, and
Witten’s spectral-action contributions remains open.
Future-work obligations: prove or refute each conjectural connection at theorem grade.

f.

Quantum-Classical Mass Scale

Theorem AH.172 (Quantum-classical macroscopic-superposition threshold: a conjecture outside DFD’s claims).
DFD predicts that quantum coherence is sustained at all
mesoscopic mass scales currently accessible by matterwave or nano-mechanical interferometry.
Proof. By the structural-reduction-of-measurement theorem (Theorem AH.162), the optical metric n = eψ is
uniquely sourced; no metric branching occurs. Decoherence in the position basis follows from the standard
ρ-coupling at rate ΓDFD ∼ Gm2 /(ℏd) · F (d/λdB ). For
an object of mass m coupled via N internal microsector
modes (with N = Ω = 60), the macroscopic-superposition
threshold corresponds to ΓDFD ·τexp ≈ 1 at experimentally
relevant times τexp ∼ 1 s.
Implication:
DFD predicts no spontaneous
macroscopic-superposition collapse at any currently
testable scale. This contrasts sharply with Diosi–
Penrose collapse (predicting collapse at τ ∼ ℏ/Egrav
for m ≳ 10−13 kg). MAQRO is the decisive test (cf.
Theorem AH.162).
g.

Mathematical-Structure Connections (Riemann,
Moonshine, Triality)

Proposition AH.173 (Riemann zeta connection at
Ω = 60, conjectural). The DFD primed-determinant ratio
at kmax = 60, formally
′

′
det(D2 )|DFD = exp[−ζD
2 (0)],

is connected to ζRiemann (s) via the Voros–Cartier framework: the trivial zeros s = −2, −4, −6, . . . correspond
to the heat-kernel coefficients a2 , a4 , a6 , . . . of the master action. Whether the non-trivial Riemann zeros at
s = 1/2 + iEn have a DFD-spectral interpretation (Berry–
Keating Hamiltonian conjecture) is open.

h.

Audit of Existing Data

Five observational streams compared with
DFD’s values. This is not a discovery claim; specific
measurements within the 2026–2030 window can decide
each case:
1. NA62 (2016–2022) K+ → π + ν ν̄ = (13.0+3.3
−3.0 ) ×
−11
10 , above DFD’s allowed window [6, 11] × 10−11
(Theorem AH.121) at its central value.
2. HERA Phase I 21-cm power-spectrum limits consistent with standard heating, as DFD predicts (its
black holes radiate at Hawking’s rate) — a consistency check, not a near-term discriminator.
3. FLAG 2024 lattice |⟨q̄q⟩|1/3 = 272 ± 5 MeV (0σ
DFD per Theorem AH.132).
4. ATLAS −1.2 < κλ < 7.2 (95% CL), consistent with
DFD’s κλ = 1, which is the Standard Model’s value.
5. NuFIT 6.0 mass splittings χ2 (2) = 0.021 against
DFD predictions at zero continuous parameters (one
data-selected binary branch choice; App. X, branchselection remark).

12.

Mathematical-Foundation Closures and
High-Precision Predictions

This subsection presents nine results on remaining structural gaps in the program, each with its grade: the
Atiyah–LeBrun–Witten triality on the DFD slice, the
Wheeler–DeWitt closed-form ground state, structural
reduction of the measurement problem, the eight-layer
single-vacuum rigidity, the photon-sphere bulk-boundary

50
partition-function identity, the A5 -equivariant sharpening of the QECC reading (which keeps it conditional
and partially established: the [[60, 3, ≥ 4]] Pauli-stabilizer
structure holds, but the literal-Hamiltonian realization remains unverified), the gen-1 fermion mass refinement, the
sphaleron-rate prefactor, and the PTA amplitudes against
the supermassive-binary background. Three named structural negative results (the Riemann no-go, the Moonshine
no-equivalence, and the sphaleron κSM measurement bottleneck) are recorded alongside.

a.

Atiyah–LeBrun–Witten Triality

Theorem AH.176 (ALW triality on the DFD slice, conditional). There exist explicit functors F1 : SpecDFD →
TwDFD (spectral → twistor), F2 : TwDFD → CatDFD
(twistor → Reshetikhin–Turaev modular tensor category),
F3 : CatDFD → SpecDFD (categorical → Frobenius algebra
→ spectral data), such that the triality identity
F3 ◦ F 2 ◦ F 1 ∼
= IdSpec
DFD

holds at the strict-fixed-point level on the DFD slice (the
spectral data over CP 2 × S 3 with χtop (CP 2 , E) = 60,
Theorem AH.77).
Proof. F1 is the composition: Connes reconstruction →
LeBrun twistor functor (using the anti-self-dual structure of CP 2 ) → Spinc -determinant lift to SU(3)/T 2 →
O(d)-bundle restriction → real-form σ. F2 is the composition: holomorphic Chern–Simons at level
p 58 on twistor
space → partition function on S 3 ∼
2/60 sin(π/60)
→ the Reshetikhin–Turaev modular tensor category
Repq (SU (2)58 ) at q = e2πi/60 . F3 is the composition:
Fuchs–Runkel–Schweigert reconstruction of a Frobenius
algebra from the MTC → modular generator DC with
eigenvalues a(2a + 3)/120 → FS-indicator real structure
→ spin-parity grading → spectral data.
The strict-fixed-point property uses the Borel–Weil–
Bott canonical isomorphism: H 0 (CP 2 , O(9)) ∼
= V(9,0) of
SU (3), which descends to SU (2)-representations matching the level-58 truncation pointwise. The composition
F3 ◦ F2 ◦ F1 on SpecDFD returns the same spectral data by
direct computation via Atiyah–Singer at F1 , the Witten
Chern–Simons identity at F2 , and Reshetikhin–Turaev
surgery at F3 .
Status. The triality is conditional: its F1 leg is the
conditional twistor theorem (Theorem AH.230), and a rigorous triality theorem remains open (Proposition AH.175).
Uniqueness across broader slices (e.g., for arbitrary Spinc
Kähler manifolds with χ(Mc ) = 3) is open. The technical
core at the F2 stage (BV-quantized factorization-algebra
rigor for holomorphic Chern–Simons) follows Costello’s
framework.

b.

Wheeler–DeWitt Closed-Form Ground State, Derivation
Status

Theorem AH.177 (Wheeler–DeWitt ground state in
DFD). The functional Ψ0 [ψ, Ξ] = N exp(−SE [ψ, Ξ]/ℏ)
on the DFD configuration space CDFD = H s (R3 ; R) ⊕
H s (R3 ; C60 ) (with internal manifold X = CP 2 × S 3 and
twisting bundle E = O(9) ⊕ O⊕5 , χ(X, E) = 60) satisfies
the Wheeler–DeWitt constraint HWD Ψ0 = E0 Ψ0 with
eigenvalue E0 = α57 MP4 = 1.59 × 10−122 MP4 . (Scope of
the match: the agreement with the observed cosmological
constant is at the exponent level, 10−122.9 — E0 itself is
∼ 14× the observed ρΛ /ρP . The physical clock dictionary
is (H0 tP )2 = Cα57 , where the exponent is derived but
C = 0.87784 ± 0.00266 is measured; it gives the early
clock H0early = 67.54 ± 0.10 km s−1 Mpc−1 , and the crosschannel value is 67.52 ± 0.09; bound-matter clocks read
C = 1, H0loc = 72.09, under the two-clock postulate;
see Apps. AP/ZS. Under the lock of the lapse to the
scale factor an eigenvalue of this Hamiltonian acts as a
curvature term, not as vacuum energy, so α57 MP4 can be
the dark energy only as a vacuum term of the action.)
Proof. The DFD Wheeler–DeWitt constraint operator
is HWD = −ℏ2 δ 2 /δψ 2 + V [ψ, Ξ] with V [ψ, Ξ] the
spectral-action potential on CP 2 × S 3 . The Euclidean
action SE [ψ, Ξ] is positive, coercive, and convex on
CDFD (via monotonicity of the spectral-action potential
and subcritical growth). Functional differentiation of
Ψ0 gives δΨ0 /δψ = −Ψ0 δSE /δψ/ℏ and δ 2 Ψ0 /δψ 2 =
Ψ0 [(δSE /δψ)2 /ℏ2 − δ 2 SE /δψ 2 /ℏ]. Substituting into
the WD equation and using the Hamilton–Jacobi identity (δSE /δψ)2 = V [ψ, Ξ] − E0 on-shell yields the
eigenvalue equation. Uniqueness follows from Faris–
Lavine (self-adjointness of −∆ + V for coercive V )
and Perron–Frobenius (positivity of the ground state).
The eigenvalue is extracted by primed-determinant ratio at gauge-normalized microsector coupling g = α
(so that det′ (g∆) = g 57 det′ ∆ = α57 det′ ∆): E0 =
det′ (D2 )|DFD /MP4 = α57 MP4 .
Status note. The printed substitution drops a 
term:
differentiating Ψ0 twice gives −ℏ2 δ 2 Ψ0 /δψ 2 = Ψ0 −
(δSE /δψ)2 + ℏ δ 2 SE /δψ 2 , so HWD Ψ0 = E0 Ψ0 requires
ℏ δ 2 SE /δψ 2 to be a ψ-independent constant; the proof
never shows this. The eigenvalue E0 = α57 MP4 via the
primed-determinant ratio is likewise asserted, not derived: det′ (D2 )|DFD is never computed above.
CMB consequences. The dictionary reading, a
Bunch–Davies vacuum on Ψ0 with slow-roll parameters
ϵW = α/(4π) and ηW = 0 and a microsector loop correction with N⋆ ≈ 27.7, gives ns = 1−28.7α/(2π) = 0.96667,
which is 1 − 2/60, since 28.7 α/2π = 2/60 to the printed
digits, and As ≈ 2.10 × 10−9 . DFD has no mechanism
for either. The tilt is a stated number; the mechanism
is open. The amplitude As is calibrated, with 32πα5
a numerical match (main volume, App. FK). The tilt
lies within 0.5σ of the CMB alone, and the combination of Planck, ACT DR6, CMB lensing and DESI gives

51
0.974 ± 0.003, about 2.4σ above it. On these premises
the tensor ratio would be 16ϵW = 4α/π ≈ 9.3 × 10−3 ; the
main volume does not adopt that dictionary (App. SM),
and the χ relic bounds the primordial B-mode ratio at
r ≲ 10−7 , so 4α/π is not a B-mode prediction.
Status of As (a calibrated amplitude). The value
As ≈ 2.10 × 10−9 above is the de Sitter dictionary result As = H⋆2 /(8π 2 ϵW M̄P2 ) = 32πα5 = 2.080 × 10−9
(−0.82σ vs Planck), with ϵW = α/(4π) and the primordial (Wheeler–DeWitt) scale on the locked Majorana rung
MR = MP α3 . The exponent α5 (the squared α3 rung)
holds only inside the dictionary at H⋆ , where no DFD
stage freezes perturbations (main volume, App. FK), and
the coefficient 32π is not first-principles derived — it
rests on the H⋆ –MR normalization, for which the corpus
supplies no de Sitter horizon, internal-volume (S 3 /CP 2 ),
or spectral-action derivation. A Planck-mass subtlety is
involved: MR = 4.74 × 1012 GeV uses the non-reduced
MP = 1.22 × 1019 √
GeV, while the formula carries the
reduced M̄P = MP / 8π; under consistent
conventions
√
the required ratio √
is H⋆ /MR = 8π (mixing the two
masses gives H⋆ = 8π MR ⇒ As = 8.3 × 10−11 , a spurious “−67σ”, and setting H⋆ = 8π MR double-counts
the conversion). As is therefore one calibrated amplitude,
and 32πα5 a numerical match; see App. AT (“new predictions”). Relatedly, the printed ηW = 0 with N⋆ ≈ 27.7 is
a mislabel: 1 − 28.7 α/2π is the slow-roll form 1 − 2/Ne
with Ne = 60, and DFD has no potential from which to
compute a slow-roll parameter (main volume, App. FK),
so the tilt is a stated number; the mechanism is open.
Falsifier. The tilt ns = 1 − 2/60 = 0.96667 (a stated
number; the mechanism is open), with no running, and no
detectable primordial tensors (r ≲ 10−7 ) is the joint observational test for the Simons Observatory and LiteBIRD:
a tilt outside 0.957–0.977 at the Simons Observatory’s
precision, or a detected primordial tensor signal, falsifies
it (main volume, App. FK and App. SM). The value
equil
fN
L = 0.007 of the de Sitter dictionary has no printed
source and is not a prediction.

c.

Quantum Measurement Problem, Structural Reduction

Theorem AH.178 (Structural reduction of the measurement problem, rigorous). The DFD elliptic optical-metric
PDE


∇· µ(|∇ψ|/a⋆ )∇ψ = −κ ⟨Ψ|ρ̂|Ψ⟩
admits a unique weak solution when µ(x) x is strictly
monotone; existence is by Browder–Minty (main volume,
App. U). There is no metric branching; the position basis
is einselected as the DFD pointer basis. The Born-rule
functional form Tr(ρP ) on Hmicro (dim = 60 ≥ 3) follows
from Gleason’s theorem once a complex inner product is
supplied—which DFD obtains from the compatible complex structure J⋆ (h = g + i ω), unique given the Hessian
positivity that the main volume’s App. QM argues but

does not prove; Gleason itself cannot select the inner
product. The Born value |⟨ϕ|Ψ⟩|2 then follows from envariance/typicality, conditional on J⋆ .
Proof. Existence of weak solutions follows from the
Browder–Minty theorem, uniqueness from the strict mono1,α
tonicity of µ(x) x, and Cloc
regularity from quasilinear
elliptic theory (main volume, App. U). No-branching: the
optical metric is uniquely determined by the c-number
expectation ⟨Ψ|ρ̂|Ψ⟩, and any branching of Ψ into orthogonal components contributes additively to the c-number
source, sourcing a single (averaged) optical metric. The
pointer basis is position because the ρ-coupling is local in
position space.
Decoherence rates: ΓDFD = ξDFD m2 d2 F (d/λdB ) with
ξDFD from the massless elliptic Green’s function of the
optical metric (Appendix A; the optical ψ is long-range
MOND, not Yukawa-screened). The smallness reflects
the c-number no-branching suppression of the paragraph above (no per-branch self-force), not a ψ mass.
Numerical (dominant c-number reading): single atom
∼ 10−43 s−1 (stable on cosmological timescales); C60
fullerene ∼ 10−38 s−1 ; MAQRO test mass (10−13 kg)
∼ 10−20 s−1 (sustained interference for 100 s); macro-cat
(1023 atoms) ∼ 1 s−1 .
Born rule, status: Gleason’s theorem fixes only the
functional form of the probability measure on the projection lattice L(Hmicro ) (dim = 60 ≥ 3) given a complex
inner product; it does not derive Born from scratch, since
it presupposes the very complex inner product / projection lattice in question (the same 60 real dimensions
admit different positive-definite inner products giving different “Gleason” weights for one ray). DFD supplies the
inner product from the compatible complex structure J⋆
(h = g + i ω), unique given the argued Hessian positivity,
and the Born value |⟨ϕ|Ψ⟩|2 then follows from envariance/typicality, conditional on J⋆ (consistent with the
Born ledger of the QM-status appendix, App. QM).
Falsifier. MAQRO experiment at 10−13 kg test mass
with 100 s coherence time decisively distinguishes DFD
(τdec ∼ 1020 s, sustained interference) from Diosi–Penrose
(τdec ∼ 10−3 s, collapse). The sustained-interference
prediction is fixed by the no-branching theorem (c-number
sourcing, no per-branch self-force), consistent with the
open-problems section; the deep-MOND-enhanced perbranch collapse alternative is excluded by the two-body
structure of the derived vertex (main volume, App. ZU).

d.

Single-Vacuum Theorem, Rigidity Layers

Theorem AH.179 (Eight-layer single-vacuum rigidity).
The DFD vacuum moduli space VDFD / ∼ (where ∼ is
the closure of gauge-equivalence, discrete-symmetry equivalence, and spectral-data equivalence) has dim = 0 and
cardinality 1. The uniqueness traces to eight rigidity statements; layer 1 is the manifold theorem, layer 6 conditional
and layer 7 a stated premise (see the note below).

52
Proof. The eight rigidity layers, with the deepest atomic
fact χtop (CP 2 ) = 3 at the foundation:
1. Manifold uniqueness CP 2 × S 3 (Appendix AB, Theorem AB.1).
2. kmax = 60 forced by χ(CP 2 , E) = 60 via Hirzebruch–Riemann–Roch.

it once the manifold is fixed — but the independence
count and the logical base must be stated as above, not
as “eight independent layers” resting on nothing prior.
Comparison: string theory (hundreds of continuous
moduli and ∼10500 flux choices, dim > 0); M-theory (G2
moduli, dim > 0); LQG (non-unique kinematic vacua);
SM EFT (26 continuous parameters); DFD (dim = 0,
|V / ∼ | = 1).

3. Flux quantum q1 = 3, derived from the isotropy
charge 6Y (main volume, App. F, the remark on
the flux quantum); the allowed values are 3λ with
λ odd.
4. Bundle decomposition (a, n) = (9, 5) forced by
minimal-padding.
5. Spinc structure unique on CP 2 (by w2 (CP 2 ) = c1
mod 2).
6. Real KO-dim 4 spectral triple unique up to isomorphism (Connes’ axioms), conditional on a carrier
that meets order zero, which C60 does not (Theorem AH.77).
7. No-go for continuous deformation D → D + δD
at Λ⋆ fixed (preserved spectral data force isolated
points). This layer is a stated premise, not a consequence of the spectral action: it holds DF at its
printed spectrum and admits as physical only the
inner fluctuations of End(E), which carry the gauge
fields (their group is selected, not forced: Theorem AH.214, Remark AH.215), not those of M60 (C)
(Theorem AH.77), which would make the 60 eigenvalues dynamical; the 57-level spectrum is not a
stationary point of the spectral action, whose stationary spectra have one nonzero level inside the
plateau and at most two beyond it for the cutoffs
tested (main volume, App. FK).
8. Cardinality 1 in the connected component (unique
limit point in the moduli of bundles satisfying 1–7).
Each layer except layer 6, conditional as stated, and
layer 7, a stated premise, is a theorem-grade rigidity
statement; the conjunction forces dim(VDFD / ∼) = 0 and
|VDFD / ∼ | = 1. The deepest atomic fact driving the
chain is χtop (CP 2 ) = 3 (identified with the discrete input
Ngen = 3; the derived q1 = 3 then propagates through
the bundle decomposition; status remarks: App. F, on
the generation count and q1 ).
Note: layer 1 is not independent. Layer 1 as printed is
Theorem AB.1, the manifold-uniqueness theorem itself,
so invoking it as one of eight “independent” rigidity layers
supporting single-vacuum rigidity is circular. This theorem has seven independent layers (2–8), each conditional
on the manifold CP 2 × S 3 having already been fixed by
Theorem AB.1, which is itself conditional on the p = 1
lens-space selection recorded in the uniqueness appendix
of the main volume. The conclusion dim(VDFD / ∼) = 0,
|VDFD / ∼ | = 1 is unaffected — layers 2–8 alone force

e.

AdS/CFT-without-AdS Bulk-Boundary Identity

Theorem AH.180 (Photon-sphere bulk-boundary partition function identity, conjectural). For any compact
mass M in DFD with photon sphere at the areal radius 3GM/c2 , the bulk path integral with Dirichlet data
∂
(gij
, ∂ϕ) on ∂Brph equals the SU (2)58 WZW boundary
partition function:
SU (2)58

∂
Zbulk [gij
, ∂ϕ] = Z∂

[J(σ) = ∂ϕ(σ)]

+
to all orders in the asymptotic expansion as r → rph
. The
boundary central charge is c = 3 · 58/(58 + 2) = 174/60 =
2.9. The HKLL-style reconstruction kernel is valid outside
the event horizon (not just near rph ).

Proof. The bulk on-shell expansion in ρ = r − rph admits proper holographic renormalization counterterms.
The boundary SU (2)58 WZW partition function uses
the affine-Lie-algebra Sugawara construction with characters χa (τ ) and modular S-matrix at k = 58. In
the order-by-order partition-function identity the bulk
Knizhnik–Zamolodchikov equations emerge from DFD
bulk equations expanded near rph . Conformal weights
∆a = a(a + 1)/60 derive from Sugawara at level k = 58;
primary fields a ∈ {0, 1/2, 1, . . . , 29} tabulate explicitly. The HKLL reconstruction kernel K DFD (r, σ; σ ′ ) =
(n(r)/n(rph ))−∆a Pa (cos γ) uses hypergeometric radial
profiles ga (r) = (rph /r)∆+ (a) · 2 F1 (. . .). Global validity for
r > 0 uses piecewise definition with analytic continuation
across the photon sphere.

f.

QECC (partially established): Pauli-Stabilizer Structure,
Literal-Hamiltonian Realization Conditional

Theorem AH.181 ([[60, 3, ≥ 4]] Pauli-stabilizer structure; literal-Hamiltonian QECC conditional). The DFD
microsector Hilbert space Hmicro = C60 carries an A5 equivariant Pauli-stabilizer structure with parameters
3
[[60, 3, ≥ 4]]: stabilizer group SDFD = ⟨Z60
⟩ ⊂ HW(60),
|SDFD | = 20, whose +1-eigenspace is the 3-dimensional
code subspace C = span{|0⟩, |20⟩, |40⟩}, identifiable with
the 3 SM generations as logical qutrits; logical Pauli operators (X̄, Ȳ , Z̄) form the qutrit Pauli group on C, with
code distance d ≥ 4 in the SM-gauge-quotient metric.
The [[60, 3, ≥ 4]] stabilizer structure holds; the literalHamiltonian QECC realization is conditional (partially
established, consistent with the assessment in App. AH 8 n

53
of the CKM/τ companion): [DF , SDFD ] ̸= 0 as operators and the code distance is 1 in the standard displacement metric (caveat to Theorem AH.144), so the literalHamiltonian and Knill–Laflamme readings are not yet
established.
Proof. The Heisenberg–Weyl group HW(60) has 3600 elements with generators X60 , Z60 satisfying X60 Z60 =
e2πi/60 Z60 X60 . The Spinc Dirac operator DF on CP 2 ×S 3
is A5 -equivariant (since the binary icosahedral 2I =
SL(2, F5 ) acts on CP 2 × S 3 by isometries). Character
theory of A5 decomposes HF = C[A5 ] into five characterisotypic components, of dimensions 1, 9, 9, 16 and 25
(the squares of the irreducible dimensions 1, 3, 3, 4, 5); the
identification of the channel space with C[A5 ] is one of
the stated structural postulates (main volume, App. Z).
The three-dimensional code space comes instead from
3
the choice of stabiliser: the eigenspaces of Z60
on C60
are twenty spaces L
of dimension 3, so the commutant of
19
3
SDFD = ⟨Z60
⟩ is
j=0 M3 (C), with |SDFD | = 20; the
3
dimension 3 is gcd(3, 60), set by the choice of Z60
.
3
Code subspace C
=
ker(Z60
− ⊮)
=
span{|0⟩, |20⟩, |40⟩} ∼
= C3 = 3 generations. Logical
X̄ permutes basis |0⟩ → |20⟩ → |40⟩ → |0⟩; logical Z̄
acts as |0⟩ → |0⟩, |20⟩ → e2πi/3 |20⟩, |40⟩ → e4πi/3 |40⟩.
Together they satisfy Z̄ X̄ = ζ X̄ Z̄ with ζ = e2πi/3 , the
qutrit Pauli relation.
Code distance d ≥√4 in the SM-gauge-quotient metric
(via the S3 winding 60 ≈ 7.75 in Berezin–Toeplitz lattice units). The [[60, 3, ≥ 4]] stabilizer structure holds;
the literal-Hamiltonian QECC realization (and hence the
Knill–Laflamme verification for error sets up to weight
3, with the quoted threshold pth ≈ 0.022) is conditional,
because [DF , SDFD ] ̸= 0 as operators and the code distance is 1 in the standard displacement metric (caveat
to Theorem AH.144). Pending that, the conservative
reading is “protected 3-dimensional subspace within a
60-dimensional Hilbert space; QECC-adjacent” (partially
established).
g.

Light-Fermion Generation-1 Mass Refinement

Theorem AH.182 (Sector-specific Toeplitz priming for
gen-1 fermion masses). The 3.32% electron-mass residual
is closed to 0.0091% via the sector-specific finite-d Toeplitz
priming factor Fe = 30/31. The same mechanism, with
sector-specific Toeplitz dimensions d, applies to up- and
down-quark masses with Fu = 45/44 (inverted direction,
conjugate-Higgs) and Fd = 57/58 (positive direction). The
mean absolute error across nine charged fermions drops
from 1.42% (leading order) to 0.61%.
Proof. The Toeplitz heat-kernel expansion at finite dimension d contributes a sub-leading 1/d correction to the
(f )
leading-order mass formula mf = MP αkf Fgeom . The
direction (multiplicative or divisive) is fixed by whether
the sector couples to the Higgs in the direct or conjugate
orientation. For the lepton (direct-Higgs) sector with

de = kmax /2 + 1 = 31 (chirality-projected half-spectrum +
null channel), Fe = (de −1)/de = 30/31. For the up-quark
(conjugate-Higgs) sector with du = a · n = 9 · 5 = 45,
Fu = du /(du − 1) = 45/44. For the down-quark
(direct-Higgs) sector with dd = (kmax − Ngen ) + 1 = 58,
Fd = (dd − 1)/dd = 57/58.
Numerical evaluation:
= 0.5109523 MeV vs. CODATA 0.5109989
• mDFD
e
MeV → −0.0091%.
= 2.1599 MeV vs. FLAG 2024 2.16 ± 0.07 →
• mDFD
u
−0.005% (within 1σ).
• mDFD
= 4.6699 MeV vs. FLAG 2024 4.70 ± 0.05 →
d
−0.640% (within 1σ).
The same {30/31, 45/44, 57/58, 14/13} priming family applies to the four primed sectors {e, u, d, ν}, with Fν =
14/13 producing the sub-meV neutrino spectrum (Theorem AH.168). Heavier-fermion stability: no priming
applies at gen-2/gen-3 because the 1/d mechanism enters
only at maximum geodesic distance nf = 2.5 from the
gen-1 Higgs vertex.
Cross-cutting consistency. The priming family is
consistent across the four primed sectors {e, u, d, ν}.

h.

Sphaleron-Rate Prefactor

Theorem AH.183 (Sphaleron-rate prefactor: structure,
not value). The DFD electroweak sphaleron-rate prefactor
is
κDFD = κSM · [1 + δB-T ],
where δB-T receives a UV contribution δUV = (T /MP )2 ∼
10−32 (structurally negligible, modes above kmax = 60
absent), a boundary contribution ∝ αW a3 from the a3
Berezin–Toeplitz heat-kernel coefficient on CP 2 × S 3 , and
a generation contribution ∝ αW ln(MR /MP )/(4π), which
is negative for MR = MP α3 . The numerical value of
δB-T is an open item: the two contributions summed
to δB-T = +0.067 ± 0.04 require mutually inconsistent
values of αW (1/14.5 vs 1/117.5) and the second carries
a sign error, so that sum does not hold. With δB-T = 0,
κDFD = κSM = 16.5 ± 2.0 (D’Onofrio–Rummukainen–
Tranberg 2014 + Bödeker–Moore 2000). The lattice κSM
uncertainty is the binding bottleneck per the structural
negative result of §AH 12 j. The result here is the structure of κDFD , not its value; ηB is a measured input here
because the thermal-leptogenesis route is closed — the
real determinant-protected lepton kernel forces its CP
source εi = 0 (baryogenesis no-go, §AH 4 b). The internal axial Berry-holonomy channel separately derives the
α4 power of |ηB |; its bias current is B−L, its printed
mechanism gives about (2–3) × 10−9 , three to five times
Planck’s value, and its sign rests on an adopted axiom
(Thm. AH.33).

54
Proof. The Klinkhamer–Manton sphaleron solution lives
in the SU (2)L × U (1)Y + Higgs sector at the EWPT.
With DFD’s optical metric n = eψ , the static-saddle
1/2
equations modify Esph by ζ = ⟨eψ ⟩T . At Tc = 159.5
GeV (matching lattice SM crossover), |ζ − 1| ≲ 10−32 ,
structurally negligible.
The prefactor decomposition δB-T = δUV + δboundary +
δgen :
• δUV = (T /MP )2 ∼ 10−32 (UV modes above kmax =
60 absent in DFD).
• δboundary = (3/2) αW a3 with a3 = 9/(8π) from
HRR on bundle E over CP 2 × S 3 : at the physical
αW (Tc ) ≈ 1/29.6 this gives +0.018, not +0.037
(which requires αW = 1/14.5).
• δgen = 3αW ln(MR /MP )/(4π) with MR = MP α3 :
since ln(MR /MP ) = 3 ln α < 0, this term is negative (−0.119 at αW = 1/29.6), not +0.030 (which
requires a negative αW , i.e. a sign error).
The sum δB-T = 0.067 ± 0.04 does not hold: no single
αW reproduces the two contributions +0.037 and +0.030.
The structure stands; the value is open.

i.

PTA Amplitudes and the Supermassive-Binary
Background

Theorem AH.184 (PTA amplitudes and the supermassive-binary A GWB). At γ = 13/3 the NANOGrav
−15
15-yr amplitude A = 2.4+0.7
agrees with the
−0.6 × 10
supermassive-binary value AGWB = (2.40 ± 0.30) × 10−15
from the SMBH binary mass function. The prediction
is the standard supermassive-binary estimate; it does not
discriminate DFD from ΛCDM astrophysics.
Proof. The DFD AGWB prediction follows from: (i)
SMBH binary mass function (Schechter fit at z = 2
with M⋆ = 6.3 × 10R9 M⊙ ); (ii) Phinney’s circular-inspiral
formula A2GWB ∝ dM n(M ) M 5/3 . Result: ADFD
GWB =
2.40 × 10−15 at 1-yr period.
The five PTA amplitudes are: NANOGrav 15-yr A =
2.4+0.7
−0.6 ; EPTA DR2new A = 2.5 ± 0.7; PPTA DR3 A =
+1.2
2.04+0.25
−0.22 ; IPTA DR2 A = 2.8−0.8 ; CPTA DR1 log10 A =
+1.0
−14.4−2.8 for a free spectral index (the other four at
γ = 13/3). Spectral index γ = 13/3 = 4.33 matches DFD;
cosmic strings predict γ = 5; FOPT γ = 4–5.
Falsifier. SKA-PTA by 2032 measures γ to ±0.02.
If γ persistently lies outside [4.0, 4.6], DFD’s nanohertz
theorem is structurally falsified.
j.

Structural Negative Results

The following three results are recorded as antitheorems: structural negative findings, recorded as the
TPI and CKM results are (Remarks AH.109 and AH.91).

a. Anti-Theorem (Riemann Hypothesis no-go). Selfadjointness of the DFD Hamiltonian HDFD on Hmicro does
not imply the Riemann Hypothesis. Three independent
obstructions: (i) finite-rank self-adjointness ̸= infinite-dim
self-adjointness; (ii) the Ω = 60 axiom forbids the kmax →
∞ limit needed for genuine Hilbert–Pólya; (iii) density
agreement ̸= pointwise eigenvalue equality (residuals up
to ±0.09 visible). Positive partial result preserved: the
Weyl counting density of {En } matches Riemann’s Nζ (E)
exactly including the 7/8 Maslov phase, and the trivialzeros / heat-kernel coefficient correspondence is rigorous.
b. Anti-Theorem (Moonshine no-equivalence). The
b8 → Leech → Monster does
chain Ω = 60 → A5 → E
not lift to a categorical equivalence. rank(C(SU (2)58 )) =
59 ̸= 1 = rank(Mod(V ♮ )) since V ♮ is holomorphic. No
braided monoidal equivalence possible, no faithful tensor
functor possible. The chain breaks at the first link (Ω =
60 ↔ A5 ), which is purely cardinality-matching. Positive
partial result preserved: A5 ⊂ Monster via Wilson 1986
contributing T5A (τ ) Hauptmodul through prime 5 of A5
— but this distinguishes 5, not 60.
c. Anti-Theorem (Sphaleron κSM bottleneck). The
DFD sphaleron-rate prediction is bottlenecked by lattice
κSM uncertainty (currently ±2.0 on κSM = 16.5). Sub1% lattice precision is required for a sub-1% test of the
Berezin–Toeplitz boundary correction; current state-ofthe-art is 12%. The bottleneck is the lattice, not DFDinternal phenomenology (Theorem AH.183).

13. Origin, Multi-Route Electromagnetism,
Quantum Gravity, and Materials-Science Predictions

This subsection presents fifteen results covering the
deepest origin questions, multi-route derivations of electromagnetism, the non-perturbative quantum-gravity path
integral plus S-matrix and the observable metric, the
three-layer back-reaction system, and materials-science
predictions. One structural negative result is recorded:
the “five proofs of Ω = 60” are four routes to the number
60.

a.

Origin of the Universe in DFD

Theorem AH.185 (Wheeler–DeWitt ground state as
the cosmogenic state). The Wheeler–DeWitt ground state
Ψ0 [ψ, Ξ] = N exp(−SE [ψ, Ξ]/ℏ) on the DFD configuration space is the proposed cosmogenic solution to the master constraint HWD Ψ0 = E0 Ψ0 with E0 = α57 MP4 (read
as in the scope note of Theorem AH.177). The start state
rests on two stated inputs, the no-boundary state with
nothing inserted and the smooth beginning, which carries
the arrow of time (main volume, App. Q). The cosmological arrow of time is welded to the dynamical complex
structure J⋆ via the optical arrow ψ̇0 > 0 (sub-statement
(ii)), not to the static CP 2 Kähler structure. The global

55
state does not dissolve the horizon problem: it fixes the
ψ fluctuation sector and the internal S 3 , not the plasma
(sub-statement (iii)).
Proof. The cosmogenic claim follows from three substatements:
(i) No-boundary condition. The DFD configuration
space CDFD has no preferred temporal boundary; Ψ0
is defined globally without specifying initial data on a
Cauchy slice.
(ii) T-asymmetry from the optical arrow on
the dynamical J⋆ , not the static CP 2 orientation. The arrow is not carried by the static internal
Kähler structure. On CP 2 , which has even complex
dimension n = 2, the flip J → −J leaves the canonical orientation vol = ω n /n! → (−1)n vol = +vol unchanged, so the i = ±J bit is orientation-invisible (the
same even-dimension fact by which CP conjugation on
CP 2 is orientation-preserving, App. L); the static CP 2
J therefore cannot break T . The asymmetry instead
arises dynamically. Time reversal T is antiunitary, so
T J⋆ T −1 = −J⋆ acts on the dynamical, field-level positivefrequency complex structure J⋆ (the global quantum i,
Theorem QM.10, App. QM), not on the static internal
J. The sign of J⋆ —equivalently the direction of iℏ∂t —is
then fixed by the optical arrow ψ̇0 > 0 (Theorem Q.10,
“Branch selection by the optical arrow,” App. Q): the
primordial fast-light boundary condition ψ0 (0+ ) → −∞
selects the future direction and hence i = +J⋆ , while
H ≥ 0—argued from the stable ψ + χ + hTT Hessian
(Theorem QM.10)—makes the bounded-below generator
A = J⋆ H/ℏ unique. The asymmetry therefore sits in the
boundary data, not in Ψ0 , which is real. Grounding: the
master dynamics are themselves T -symmetric; the arrow
rests on a primordial boundary condition, the smooth
beginning of the main volume’s App. Q—the same lowentropy-past input that every cosmology requires, not a
consequence of time-symmetric dynamics (Rem. Q.11).
(iii) Horizon uniformity. The printed global state
does not supply the smoothness of the CMB at z = 1100:
Axiom PO of the main volume’s App. Q fixes the ψ
fluctuation sector and the internal S 3 , not the plasma,
and the print names no process that makes the plasma;
the smoothness is stated, in the smooth beginning (main
volume, App. Q). The pre-BBN optical engine would put
the observed sky in causal contact, but it accelerates the
observable scale factor, so it needs a source of negative
pressure and a reheating stage, which the print does not
supply (main volume, App. Q).
The exponent 57 of E0 = α57 MP4 is that of Theorem AH.158; the coefficients differ, and the observed
dark-energy density carries the measured R (main volume, App. AU). (Theorem AH.177’s own eigenvalue step
is asserted, not derived; see its status note, §AH 12 b.)

b.

Four Routes to Ω = 60

Theorem AH.186 (Four routes to Ω = 60). The four
converging derivations (HRR Euler characteristic on bundle E over CP 2 ×S 3 , |E8 roots|/4, |A5 | directly, RT MTC
at level k = 58) all give 60.
Proof. Direct equivariant Lefschetz–RR on the orbifold X = CP 2 × S 3 /2I (where 2I = SL(2, F5 ) is
the binary icosahedral
group double-covering A5 ) gives
P
χequiv (X, E) = g L(g) = |A5 | = 60 via the Lefschetz
b8
fixed-point trace. The McKay correspondence 2I ↔ E
pairs the nine irreducible representations of 2I with
the nine nodes of the affine E8 diagram, and E8 has
240 = 4 × 60 roots. The RT modular-tensor-category
at level 58 gives 60 through the Sugawara level-shift
k + h∨ = 58 + 2 (it has 59 simple objects).
a. Anti-Theorem (Five proofs of Ω are fourequivalent). The five derivations are four routes to the
number 60.
c.

Origin of CP 2 × S 3 : Eight-Layer Manifold Rigidity

Proposition AH.187 (CP 2 × S 3 is DFD-admissible
under eight rigidity axioms; uniqueness is conjectural
pending exhaustive enumeration). Among compact 7manifolds, the pair (M, E) = (CP 2 ×S 3 , O(9)⊕O⊕5 ) satisfies the eight DFD-admissibility axioms. Eight rigidity
layers narrow the candidate space substantially, and explicit refutations of named alternatives (S4 , CP 3 , S2 ×S2 ,
S7 , T4 ×S3 , K3×S3 , Calabi–Yau 3-fold, G2 holonomy, lens
spaces) each fail at least one rigidity layer. Strict uniqueness across the full 7-manifold landscape is conjectural
pending exhaustive enumeration; the present statement
is a top-down narrowing argument, not a from-below
classification.
Plausibility argument. The eight rigidity layers each force
one factor of the parameter space:
1. Compactness ⇔ finite mode count.
2. Total dimension 7 ⇔ correct gauge-group dimensional analysis (dim GSM = 12, after KK reduction).
3. Product structure Mc × Mg ⇔ Wang’s classification
of compact Lie-coset products.
4. χtop (Mc ) = 3 plus positivity and simple connectivity (main volume, App. AB) ⇒ Mc = CP 2 .
5. π3 (Mg ) = Z + free SU (2) action ⇔ Mg = S 3 .
6. Spinc structure on CP 2 ⇔ w2 = c1 mod 2.
7. Self-dual Einstein metric of positive scalar curvature
(Hitchin) ⇔ Fubini–Study.
8. Minimal-padding bundle E = O(9)⊕O⊕5 ⇔ explicit
HRR computation χ(E) = 55 + 5 = 60.

56
The eight refutations of named alternatives are individually rigorous. It is a proposition, not a theorem, because
of the distinction between (a) showing that named alternatives fail, and (b) ruling out all conceivable alternatives via
exhaustive product-decomposition + bundle-classification
enumeration. The latter requires a separate classification
result not in hand.

under continuous-parameter rates the evidence ratio is
very large, while under appropriate discrete-choice rates
(Theorem AH.207 below) the ratio is reduced by a much
smaller factor. We present the conservative accounting in
Theorem AH.207.
e.

Scope note. This statement is not a uniqueness theorem. The “ruled out alternatives” list (S4 , CP 3 , etc.)
demonstrates non-admissibility for the named candidates;
it does not constitute a from-below classification. A
uniqueness theorem awaits either an exhaustive classification of compact 7-manifolds with product structure
satisfying axioms 1–8, or a rigorous obstruction argument
blocking all unnamed candidates. Until then, the present
statement is recorded as a proposition. The main volume’s
uniqueness theorem (App. AB) is conditional on seven
stated axioms and does not supply that classification.

d.

Master Derivation Tree of the Standard-Model Constants

Theorem AH.188 (Derivation tree of the SM + Λ + H0
constants, given the stated inputs). All 34 continuous parameters of the Standard Model + Λ + H0 follow from the
pair (Ω = 60, MP ) given the discrete, prefactor and continuous inputs of App. FK of the main volume (see caveat).
The nine charged-fermion masses carry a 0.61% mean error (Theorem AH.182). Scope caveat (consistent with the
master fermion ledger, App. FK): what is derived firstprinciples is the α-power exponent ladder and the equalexponent ratio structure; the absolute charged-fermion
spectrum below the top imports seven data-selected prefactors (main volume, App. FM), and the derivation tree
reproduces the 34 values given those selected/fitted inputs
rather than from (Ω, MP ) alone. The 0.61% mean error
measures the fit quality of that structure, not full firstprinciples derivation. Bayesian model comparison against
the Standard Model (which has 26 continuous free parameters), under standard descriptive-length penalties for
continuous parameters, yields a favorable evidence ratio
once those discrete/prefactor inputs are counted; we discuss this conservatively in Theorem AH.207 below, where
the continuous-parameter and discrete-axiomatic-choice
categories are disentangled.
Proof. The 34-parameter master table lists each parameter with: (i) topological identity from Ω; (ii) numerical evaluation; (iii) PDG/CODATA comparison; (iv)
residual. Examples: α−1 = (π 3/2 /24) Tr(Y 2 ) Ω (Ω +
3)/(Ω + 4) [1 + 7/(80(Ω + 4)2 − 80)] = 137.03599985412 . . .
(Theorem AH.167); Λ ∝ α57 MP4 (Theorems AH.158
and AH.177, the latter’s eigenvalue step
√ asserted, not derived, per its status note); v = MP α8 2π = 246.09 GeV
(the Higgs hierarchy theorem); Σmν = (14/13) π MP α14 ·
(Branch-B mixing) = 61.462 meV (Theorem AH.168).
The Bayesian comparison against the Standard Model is
sensitive to how discrete-axiomatic choices are penalized;

Four Presentations of α−1

Theorem AH.189 (Four presentations of α−1 ). The
closed-form value of α−1 is presented in four ways within
the heat-kernel/spectral family:
1. Connes spectral triple: inner-automorphism
gauge group + heat-kernel a4 Maxwell action.
2. Berezin–Toeplitz quantization: line bundle L =
O(1) on CP 2 at level k = kmax = 60, the truncation,
not the level of an SU (2) Chern–Simons theory.
3. Optical-metric variational principle:
direct
R
variation of the Maxwell term − 14 d4 x n2 F 2 of
the optical metric.
4. Anomaly cancellation: ind(DE ) = χ(CP 2 ×
S 3 , E) = 60 + Burnside identity 60 = 12 + 32 +
32 + 4 2 + 5 2 .
The closed form is that of Theorem AH.167, whose last
21 ppm rest on the bracket coefficient, whose provenance is
open. The four presentations are connected by named functorial bridges (Bordemann–Meinrenken–Schlichenmaier
and Connes–Moscovici) and therefore share substantial
structural content; the categorically distinct fifth route
(Verlinde-modular Hopf invariant) is presented in Theorem AH.206 below.
Proof. Cross-presentation consistency is verified:
Spectral → Toeplitz. 
Both reduce to the heat-kernel
expansion of Tr exp −tD2 at the truncation k = kmax =
60; the Berezin–Toeplitz dictionary Tk (f ) = Πk Mf Πk
identifies them as different polarizations of the same algebra (Bordemann–Meinrenken–Schlichenmaier convergence: Tk → Dspectral as k → ∞). The two presentations
are related but not equivalent at finite k.
Toeplitz → Variational.
The Toeplitz partition
R
function
Z
[A]
=
DA
exp(−S
BT
BT [A]) with SBT =
R
(k/4π) F ∧ ⋆F reduces to the variational Maxwell action
with g 2 = 2π/k = 2π/60; the variational route inherits its
numerical normalization from the Toeplitz construction.
Variational → Anomaly. The a4 heat-kernel coefficient that produces the Maxwell action in the spectral
route and the prefactor of the chiral anomaly in the
Atiyah–Singer index are related but not equal: on CP 2
the charge-squared part of the index is q 2 /2 and that of
a4 is q 2 /24, a factor 12 apart. The anomaly-free Standard Model content, an input, gives Tr(Y 2 ) = 10 on
the three-generation (3, 2, 1) bundle (Slansky 1981 p. 96
per-generation 10/3 × 3 generations).
Anomaly
P → Spectral.PThe anomaly cancellation
conditions f Yf3 = 0 and f Yf = 0 are equivalent to

57
the algebraic conditions that the inner automorphisms of
the spectral triple form a closed gauge group; this closure
is the Connes gauge-closure axiom.
Within the heat-kernel/spectral family the four presentations are related-but-not-equivalent (functorial bridges
named above); the categorically distinct fifth route is the
Verlinde-modular Hopf-invariant presentation of Theorem AH.206.
Burnside identity (numerical verification). The
five irreducible representations
of A5 have dimensions
P
{1, 3, 3, 4, 5}, and |di |2 = 1+9+9+16+25 = 60 = |A5 |.
f.

Non-Perturbative Quantum-Gravity Path Integral

Theorem AH.190 (DFD QG partition function: the
finite microsector factor). The DFD partition function
Z
ZDFD =
Dψ Dhµν DAµ Dψ̄ Dψ d2N ϕ exp(iSDFD /ℏ),

g.

Observables and the Physical Metric

Theorem AH.191 (Observables and the physical metric). In DFD all physical observables are functions of the
physical metric g̃µν (written here in the conformal-class
shorthand e2ψ gµν ; the physical matter-coupling metric
is the non-conformal diag(−c2 e−ψ , e+ψ , e+ψ , e+ψ ) of the
main-text formalism, whose opposite-sign lapse/space scaling gives the active-mass source ε + 3P and light-bending
γ = 1). DFD keeps a preferred foliation, its absolute time
(App. WS); the statement concerns the spatial metric
only.
Proof. With the flat arena fixed, the physical metric determines ψ, so any observable computed from SDFD is a
functional of the physical metric and the preferred foliation. The lock leaves only the spatial diffeomorphisms
that preserve the arena’s volume form (App. WS, Proposition WS.6(a)). Asymptotic structure at I + : BMS group
identical to GR (Theorem AH.217).

N = 57,
on the DFD covariant configuration space Ccov
DFD has a microsector factor that is finite, with no UV regulator other
than the Berezin–Toeplitz cutoff at kmax = 60. All correlation functions ⟨ϕ1 · · · ϕn ⟩ are finite Berezin–Toeplitz
traces on Hmicro = C60 . The five-way measure factorization used in the proof holds given three corpus-written
selections (plateau cutoff, spectral-sector matter definition,
constant g = α−1 ); with those in place the micro factor is
finite; the four-dimensional gauge, graviton and matter
factors are not constructed (Theorem AH.231, scope).
Proof. The Parisi–Wu Brownian construction yields the
stationary measure dµDFD = Z −1 exp(−SE /ℏ) DΦBT as
the solution to the Stratonovich stochastic PDE; existence
and uniqueness are asserted from the Bochner–Minlos
representation theorem combined with the Osterwalder–
TT
Schrader axioms. The five-way splitting ZDFD = Zgrav
·
Zψ · Zmicro · Zgauge · Zfermion is a derived conditional lemma
(not a bare declaration): (i) the sole Grassmann pair in
the measure is the matter Dirac field of Smatter , so no
anticommuting variable multiplies the micro block (field
content of the written action); (ii) the 57 microsector
amplitudes ϕi appear in no term of SDFD coupling them to
any other sector (their own quadratic weight is the micro
block itself), so the commuting quadrature d2N ϕ is actionindependent of every other sector; (iii) given the plateau
cutoff (exact quadraticity of the micro block, App. O),
the spectral-sector definition of matter (zero modes of
DF ; the ϕi span the orthogonal nonzero eigenspaces),
and constant g = α−1 , the micro Gaussian factors out
′
exactly, contributing π 57 /(g 57 det{} K). Legs (i)–(ii) are
unconditional in the written action; leg (iii) holds given
the three named, corpus-written selections. The micro
factor is finite by the BT-cutoff bound on the Hilbertspace dimension (60-mode). Correlation functions are
computed via finite Berezin–Toeplitz quadratures.

h.

Quantum-Gravity S-Matrix

Theorem AH.192 (Unitary, finite (conjectural), CPT-invariant DFD
The DFD scattering matrix


R S-matrix).
S = T exp −i Hint dt is well-defined on the asymptotic
Hilbert spaces Hin and Hout (free Fock spaces of graviton, photon, W /Z, gluon, quarks, leptons, Higgs). The
S-matrix is unitary (S † S = 1), CPT-invariant, asymptotically Lorentz-invariant, and IR-consistent (Weinberg,
Cachazo–Strominger, Sen soft theorems all reproduced).
UV-finiteness, conjectural, is argued from the Berezin–
Toeplitz cutoff at kmax = 60, with the power-counting
bound |M(L) | ≤ const · 602L · κ2L+2 · poly(s, t, u). The
statement holds below the spectral cutoff: the Weyl-squared
term of the spectral action implies a spin-2 pole at about
3.6 reduced Planck masses for f0 = 1 (main volume,
App. WS), above the cutoff, where it does not extend.
Proof. The asymptotic completeness theorem establishes
Møller wave operators on the optical-metric asymptotic
states. Unitarity: Cutkosky cutting rules at 1- and 2loops, BRST/Kugo–Ojima quartet mechanism on Hphys .
CPT: invariance of the BT cutoff under the antiunitary
CPT operator. Asymptotic Lorentz: g̃ → η as r → ∞
via the Asymptotic Metric-Coincidence Lemma. Soft
theorems: explicit reproduction of Weinberg (leading), CS
(subleading), Sen (sub-subleading); BMS Ward identities;
gravitational memory effect. UV: the BT-cutoff powercounting bound of the statement, conjectural. Goroff–
Sagnotti tower cancels termwise.
i.

Three-Layer Back-Reaction System

Theorem AH.193 (Self-consistent classical-quantum
back-reaction in DFD). The DFD back-reaction system
is self-consistent at three layers: (i) classical ψ acts on

58
matter via the modified geodesic equation in the optical metric; (ii) classical matter sources ψ via the elliptic PDE ∇ · [µ(|∇ψ|/a⋆ )∇ψ] = −κ (ε + 3P )matter , the
Tolman active gravitational mass (the main-text Optimatter
cal Source Theorem; → κT00
in the non-relativistic
P ≪ ε limit); (iii) quantum matter sources classical
ψ via the BT-renormalized active (Tolman) expectation
κ⟨ε̂ + 3P̂ ⟩ren , obtained by varying the one-loop effective
action Γ[ψ] through the same non-conformal optical metric (the main-text Optical Source Theorem; the ⟨T̂00 ⟩-only
form is its non-relativistic limit). All three layers admit
unique solutions; the Banach contraction theorem and the
BT-positivity bound preclude pathological feedback.
Proof. Layer (i): the geodesic equation ẍµ + Γ̃µαβ ẋα ẋβ =
0 is derived from the variational principle on g̃; PPN
parameters γ = β = 1, α1 = α2 = α3 = 0 exactly
(Mercury perihelion 43”/century recovered).
Layer (ii): existence via Browder–Minty and uniqueness
from the strict monotonicity of µ(x) x (main volume,
App. U).
Layer
(iii):
the
semi-classical
equation
∇ · [µ(|∇ψ|/a⋆ )∇ψ] = −κ⟨Ψ|(ε̂ + 3P̂ )|Ψ⟩ren admits
unique solutions for any normalized quantum state via
the BT-renormalization scheme. The active-mass form is
forced by the same variational rule as the classical Optical
Source Theorem: ⟨T̂ ab ⟩ren (∂g̃ab /∂ψ) = ⟨ε̂ + 3P̂ ⟩ren is
exact for any symmetric renormalized tensor (the trace
anomaly shifts ⟨ε̂ − 3P̂ ⟩, and the stress it induces carries
a small ⟨ε̂ + 3P̂ ⟩ as well, of order H 4 in the radiation
era and about 5 × 10−25 of radiation’s share at TL
(computed), so radiation sources at 2⟨ε̂⟩ to that accuracy,
not 0 or ⟨ε̂⟩); it reduces to κ⟨T̂00 ⟩ren non-relativistically.
Banach contraction on the joint (ψ, ⟨T̂ ⟩) map ensures
self-consistency. (With the parent strain’s σ kept the
exterior’s event horizon carries Hawking’s temperature
from its surface gravity; any correction to it is written as a
correction to that base value, per App. AP, Cor. AP.36.)
Strong-field tests against PSR B1913+16, J0737−3039,
J0740+6620, EHT shadows and the LVK ringdowns
ˆ
(GW250114, δf
220 = 0.02 ± 0.02, 90% C.L.) are passed as
general relativity passes them: with the parent strain’s σ
kept these observables are general relativity’s (App. WS;
the ringdowns argued beyond first order in the spin).
j.

Near-Room-Temperature Superconductivity Prediction

Theorem AH.194 (ψ-mediated electron pairing and
Tc in Hg-1223 at 50 GPa: a conjecture outside DFD’s
claims). The ψ-mediated electron-electron interaction in
DFD adds an attractive coupling λDFD ≈ 0.20–0.50 to the
standard phonon coupling in cuprate superconductors. In
the optimal Hg-1223 composition under 50 GPa pressure,
the DFD-modified BCS gap equation predicts a critical
temperature in the range Tc ∈ [149, 283] K corresponding
to the full coupling range; under the central preferred coupling λDFD ≈ 0.40 (motivated by Fermi-surface-averaged

ψ-mode coupling at optimal cuprate doping p = 0.16), the
central prediction is Tc ≈ 240 K. The pairing symmetry
is dx2 −y2 , inherited from the CuO2 planar geometry. The
doping-dependent isotope coefficient α(p = 0.16) ≈ 0.06
matches observation.
Proof. The ψ-mediated 4-fermion vertex from the master
action gives an effective attraction between electrons via
ψ-quantum exchange:
VDFD (q) = −

e4

1

4π 2 ℏ2 c2 q 2 + m2ψ

,

where m2ψ is the effective ψ-mode mass set by lattice dispersion. Integrating over the cuprate Fermi
surface gives λDFD as a function of doping, lattice
constant, and pressure.
The weak-coupling BCS
result with the DFD-modified coupling, kB Tc =
1.13 ωD exp[−1/(λphonon + λDFD − µ∗ )], gives Tc . For
Hg-1223 at P = 50 GPa, ωD ≈ 70 meV (compressed),
λphonon ≈ 0.45, µ∗ ≈ 0.10. With λDFD in its assumed
range [0.20, 0.50], this gives Tc ∈ [149, 283] K; under the
central preferred value λDFD ≈ 0.40 (motivated by Fermisurface averaging at optimal cuprate doping), the central
prediction is Tc ≈ 240 K.
d-wave pairing follows from the C4v symmetry of the
planar lattice and the angular dependence of the ψmediated vertex; isotope coefficient and pseudogap T ∗
predicted via standard Eliashberg analysis with the DFDmodified pairing kernel.
Status. The coupling λDFD is not derived from the
action, so this is a conjecture outside DFD’s claims: an
Hg-1223 measurement at P = 50 GPa yielding Tc outside
the predicted range at > 3σ refutes the conjecture, not
the theory.

k.

ψ-Protected Topological Phase (ψPT)

Remark AH.195 (ψPT: conjectured ψ-protected topological phase — not a theorem). A theorem-grade claim
of a laboratory-accessible topological phase driven by
a gravitational-gradient shift of the Berry connection,
ABerry → ABerry + αDFD ∇ψ, with a critical gradient
quoted as ∇ψc ≈ 10−15 m−1 and a centrifuge detection
scheme fails on three counts. (i) Its own scale formula,
∇ψc ∼ αMP /kmax , evaluates to ∼ 7.5 × 1030 m−1 (with
MP the inverse Planck length), some 46 orders of magnitude above the quoted number and far beyond any
laboratory gradient, so no laboratory implementation
follows from the printed coupling. (ii) A pure-gradient
shift of a Berry connection carries zero Berry curvature
and therefore cannot by itself change a Chern class; a
curvature-generating mechanism would be required and
none is printed. (iii) As written the coupling is dimensionally inhomogeneous (a real-space gradient added to a
momentum-space connection, with αDFD glossed as the
acceleration scale a0 ). The underlying idea — that a
ψ-gradient could imprint on band topology — is retained

59
as a conjecture outside DFD’s claims only: it is not a
prediction, carries no detection scheme, and must not be
cited as a laboratory test of DFD unless a derivation supplying a curvature-generating coupling and a re-derived
threshold is printed.

l.

Metamaterial Analog Program for DFD

Theorem AH.196 (Engineered metamaterials simulate
DFD gravitational predictions). For any DFD prediction
in the gravitational/cosmological regime, an analog laboratory simulation exists in metamaterials with engineered
refractive index n(x) = eψ(x) . In particular, Schwarzschild
geometry, the photon-sphere holographic duality, and the
+0.232% sector-resolved Casimir shift (a conjecture outside DFD’s claims) can all be studied on tabletop optical
setups.
Proof. The optical-metric postulate n = eψ is mathematically identical between cosmological and laboratory contexts. Photonic-bandgap structures with 60 modes and
A5 symmetry simulate the DFD microsector [[60, 3, ≥ 4]]
code. Casimir cavities with metal/dielectric walls test
the α/π = 0.232% sector-resolved shift at sub-0.1% precision.

m.

Multiferroic Magnetoelectric Memory Effect

Theorem AH.197 (Persistent magnetoelectric coupling
from DFD optical-metric: a conjecture outside DFD’s
claims). The conjectured effect is a persistent magnetoelectric memory effect in multiferroics with magnitude
δαME ≈ 6 × 10−17 ps/m and decay timescale τψ ≈ 100 s.
Proof. The DFD-induced magnetoelectric coupling has
DFD
the form αME
(ψ) ≈ (αem /π) ψlocal · χ0P M , where ψlocal is
the local gravitational potential and χ0P M is the conventional cross-susceptibility. The decay timescale follows
from the ψ-coherence length on lattice scales. Memory
persistence: the optical-metric coupling preserves M · P
correlations over timescales much longer than typical
multiferroic decoherence (which is set by lattice phononmagnon coupling).

n.

DFD Completeness Theorem on 40 Fundamental
Questions

Theorem AH.198 (DFD’s answers to 40 fundamental
physics questions). With respect to a list of 40 fundamental physics questions Q1 –Q40 (covering origin of the
universe, dimensionality, gauge group, fermion generations, masses, Yukawa couplings, Higgs sector, Λ, H0 ,
dark energy, dark matter, black-hole information, Hawking radiation, Wheeler–DeWitt, and related topics), DFD
provides:

• 34 answers (X = 34) cited to specific results in
this appendix, each at the grade printed there (after
the partial-grade closures, the CKM apex at partial
grade, and the no-go for the optical-screen Hubble
reconciliation, Theorem AH.201);
• 1 partial-grade answer (P = 1): the CKM
apex prediction (both Path A and Path B preserved
as conjectural candidate selection-rule extensions
pending LHCb Run 4 + Belle II);
• 5 open gaps (K = 5): the underived size of
the two-clock ratio, the charged-fermion residuals
(Theorem AH.200), plus Q12 origin (§AH 13 a), Q27
(the anthropic principle) and Q28 (the string landscape). Theorems AH.223 and AH.224 answer these
two only in part, the first for the derived exponents
and ratios and the second given the rigidity premise
(layer 7) of Theorem AH.179, so they are counted
open.
X + P + K = 40 (34/1/5). No question is omitted; five
are open.
Proof. The three answer categories are mutually exclusive
and exhaustive. After the closure of Theorem AH.203,
the CKM apex at partial grade, the no-go for the opticalscreen reconciliation (Theorem AH.201) and the failure of
the charged-fermion running closure (Theorem AH.200),
the cardinality is 34/1/5 (answered/partial/open-gap).
The size of the two-clock ratio is counted as an open
gap.

o.

Single Master Equation for Origin and Evolution

Theorem AH.199 (Master equation closure: origin +
evolution + observables). The DFD master equation
SDFD = Tr f (D/Λ∗ ) + ⟨Ψ, D Ψ⟩
together with the master invariant Ω = 60 and the dimensional anchor MP organize: (i) the Wheeler–DeWitt
cosmogenic state Ψ0 ; (ii) the entire SM Lagrangian via
heat-kernel expansion; (iii) all 34 continuous parameters
(the exponent ladder and ratio structure at theorem grade;
the absolute charged-fermion spectrum below the top given
the discrete selection bits and fitted prefactor blocks of
Theorem AH.188); (iv) the four presentations of α−1 ;
(v) the microsector factor of the QG path integral; (vi)
the three-layer back-reaction system; (vii) the materialsscience predictions; (viii) the tally of the 40 fundamental
questions of Theorem AH.198.
Proof. The proof is the conjunction of Theorems AH.185–
AH.198 plus all upstream theorems referenced therein.
Each item (i)–(viii) is established by a specific theorem
with proof.

60
14.

S-matrix is unitary and CPT-invariant on the DFD
asymptotic Hilbert space (free Fock space of hT T
quanta, photons, gluons, W/Z, fermions, Higgs); its
UV-finiteness from the BT cutoff at kmax = 60 is
conjectural.

Translation Dictionary: DFD-Native Objects vs.
Graviton Language

This subsection establishes the translation map between DFD’s native objects and the graviton-language
used in standard quantum-field-theory of gravity. DFD’s
primitive variables are the scalar field ψ(x, t) on flat substrate R1,3 , the optical metric g̃µν = e2ψ ηµν (a coordinategauge shorthand for the radiative/graviton-translation
map only; the physical matter-coupling metric is the
non-conformal g̃µν = diag(−c2 e−ψ , e+ψ , e+ψ , e+ψ ) of
the main-text Mathematical Formalism section, whose
opposite-sign lapse/space exponents give the active-mass
source ε + 3P and thereby avoid the Nordström conformal
defect — the conformal form here must not be read as the
physical coupling), and the transverse-traceless radiative
sector hTijT , the spin-2 part of the parent strain whose
trace is ψ (App. WS). DFD does not have gravitons in the
standard QFT-of-gravity sense as a fundamental quantum
field; instead, the graviton-language is shorthand for a
derived radiative sector.
• Graviton field hgrav
←→ hTijT , the transverseµν
traceless component of the optical-metric perturbaTT
tion. The DFD mapping is hgrav
µν (x, t) = hij (x, t)
for spatial indices in the radiation zone, with the
trace carried by ψ and the vector and scalarlongitudinal pieces constrained, non-radiative fields
of the parent strain (App. WS).
• Soft-graviton theorems (Weinberg, Cachazo–
Strominger, Sen) ←→ TT-sector waveform corrections. The leading Weinberg theorem in DFD-native
language is the universality of the hTijT asymptoticflatness limit at scri+ ; the sub-leading and subsub-leading corrections (Theorem AH.203) are explicit TT-sector waveform predictions for BNS lateinspiral phase shifts at order α2 ≈ 5 × 10−5 , accessible at Einstein Telescope and Cosmic Explorer.
• BMS group at scri+ ←→ DFD’s optical-metric
boundary structure. The BMS4 supertranslation
algebra is realized in DFD via the asymptotic (r →
∞) limit of the optical metric: g̃ → η + hT T /r
with ψ → 0 at O(1/r). The Asymptotic MetricCoincidence Lemma (Theorem AH.217) establishes
that the residual asymptotic-symmetry group equals
the GR BMS group at all orders.
• Quantum-gravity path integral ←→ Parisi–
Wu Brownian path integral over (ψ, hTijT ) at finite
Berezin–Toeplitz cutoff ΛBT , extended to the continuum limit via Theorem AH.231. The graviton-loop
expansion of standard QFT-of-gravity corresponds
in DFD to the heat-kernel expansion of the spectral
action plus the Berezin–Toeplitz quantization of the
radiative sector.
• QG S-matrix ←→ scattering amplitudes in DFD’s
microsector ⊕ radiative sector ⊕ matter sector. The

• Trans-Planckian BH formation amplitude
/ BH echoes ←→ photon-sphere [[60, 3, ≥ 4]]
stabilizer-code projection of the radiative sector.
The third-polarization echo estimate of Theorem AH.233 does not apply: the far zone carries
the two tensor polarizations only, and gravitationalwave echoes are absent because the remnant has a
Kerr horizon (App. WS; argued beyond first order
in the spin).
• Gravitons as quantum field ←→ not a primitive
object in DFD. Wherever “graviton” appears in
the corpus, it should be read as “quantum of the
hT T radiative sector of the parent strain.” The
translation is not an additional postulate; it is a
definition.
Status of the translation. The dictionary above is
a definitional convenience. Every theorem in the corpus
that uses graviton-language admits a parallel statement in
DFD-native variables; the substantive predictions (BNS
phase shifts, soft-theorem coefficients, S-matrix unitarity)
are the same in either language. The graviton language
is kept throughout the corpus as the standard QFT-ofgravity vocabulary; the dictionary makes the translation
explicit, so the question of where gravitons come from in
a scalar theory on flat space is answered in one place.

15.

Closures of Remaining Partial-Grade Theorems

This subsection presents one theorem-grade closure
(Theorem AH.203), one modeling interpolation (Theorem AH.205), one closure that does not hold (Theorem AH.200: the charged-fermion residuals are not
running artifacts), one no-go (Theorem AH.201), and
one candidate-grade result (the CKM apex, Proposition AH.204). That proposition remains candidate-grade,
not theorem-grade.

a.

Yukawa Precision: Running Does Not Close the Residuals

Theorem AH.200 (Two-loop running does not close
the charged-fermion residuals). The four charged-fermion
mass residuals above 0.5% (mµ +2.72%, mτ +1.12%,
mb −0.83%, mt +0.78%) are not√running artifacts between the matching scale Λtop = α MP ≈ 1.04 × 1018
GeV and the PDG/FLAG reference scales. Two-loop
Standard-Model running of the masses from Λtop gives
the factors {ηµ , ητ , ηb , ηt } = {1.105, 1.094, 4.48, 2.52}, not
the {0.9735, 0.9889, 1.0084, 0.9923} that would remove the

61
residuals; those equal 1/(1 + residual), so their agreement
with the residuals is arithmetic, not running. The mean
charged-fermion error stays at the value reached after
Theorem AH.182.
Proof. No reading gives the required factors: other start
scales (MP , 2 × 1016 GeV, MR , 1010 GeV), Yukawa couplings in place of masses, inverse factors, and two-loop
over one-loop ratios all fail (one-loop running gives 1.095,
1.084, 4.27, 2.39). QED alone fixes ητ /ηµ = 0.990 at
any common scale, where 1.016 would be needed, and
Standard-Model running from any scale at or above Mt
gives ηb /ηt ≥ 1.45, where 1.016 would be needed (computed). The priming-family discipline of Theorem AH.182
cannot be extended to gen-2/gen-3: the Berezin spectral expansion shows priming entries scale as 1/dnf , depleting from ∼ 3% at gen-1 to ∼ 0.1% at gen-2 and
∼ 0.003% at gen-3. The four-element priming family
{30/31, 45/44, 57/58, 14/13} is therefore structurally complete.
Status: closing these residuals would need new structure at the matching scale, which is not printed.

b.

No optical-screen reconciliation of the Hubble clocks

Theorem AH.201 (No optical-screen reconciliation of
the Hubble clocks). Under the current DFD equations,
no homogeneous or conformal optical screen can shift the
acoustic angle or distinguish the CMB, BAO, supernova,
and time-delay clocks. The claim of a joint H0 = 72.07 ±
0.4 km/s/Mpc from such a screen therefore fails. DFD’s
acoustic clock is H0early = 67.52 ± 0.09 km/s/Mpc; the
local high-H0 measurements are read on the bound-matter
clock, H0loc = 72.09, by the two-clock postulate of the
main volume (App. ZS), a ratio between frames and not
a screen.
Proof. The exact conformal identity of Appendix ZS in
the main volume multiplies the proper sound-horizon
rod and angular-diameter distance by the same factor,
so θ∗ = rs /DA is invariant. The homogeneous background mode separately adds no screen beyond the expansion. The remaining inhomogeneous mean is 3.8 × 10−6 ,
about 3500 times below the required mean shift. The
low-redshift arithmetic of that claim is also inconsistent:
′
∆µ = 0.039 implies DL
/DL = 100.039/5 = 1.01812 and
hence 73.04/1.01812 = 71.74, not 72.07; the reverse convention gives 74.36. Finally, the four additional corrected
probe values and covariance needed for the asserted sixprobe combination are not given. The proposition itself
therefore fails, not only its explanation.
Remark AH.202 (What the theorem does and does not
exclude). The first-peak observable holds through the
measured-clock no-screen route: 100θ∗ = 1.04110 and
ℓ1 ≃ 220. Local and inhomogeneous lensing observables,
including anisotropic correlation tests and the separately

scoped domain-local screen, are not excluded by this
theorem. What is excluded is the use of a global/probeselective screen to produce one reconciled 72.07 clock.
Current test: precision bright standard sirens, which
use no calculated rate, test the two-clock postulate of the
main volume (App. ZS): its prediction is H0loc = 72.09,
and sirens settling at 67.5 falsify it.
c.

Soft-Graviton Sub-Leading Corrections

Theorem AH.203 (DFD-specific corrections to Cachazo–Strominger and Sen soft theorems). The DFDspecific corrections to standard soft-graviton theorems
are: (i) Weinberg leading S (0) has strict zero correction (universality preserved); (ii) Cachazo–Strominger
(1)
(1)
sub-leading SDFD = SCS (1 + α2 δ (1) ) with δ (1) = +1
from the Padé-(2,2) residue; (iii) Sen sub-sub-leading
(2)
(2)
SDFD = SSen (1 + α2 δ (2) ) with δ (2) = −1/2 (sign flip
from second-order Padé residue). Magnitude bounded by
α2 ≈ 5.33 × 10−5 .
Proof. The leading Weinberg theorem is universal under
the Asymptotic Metric-Coincidence (AMC) lemma used
in Theorem AH.192, which guarantees g̃ → η at scri+ at
O(1/r), and is preserved under the BT cutoff at kmax = 60
(UV cutoff, not IR). The sub-leading and sub-sub-leading
corrections come from finite-kmax heat-kernel expansion
of the spectral action; explicit Padé residue computation
gives δ (1) = +1 and δ (2) = −1/2. BNS coalescence
waveform phase-shift from S (1) correction is ∼ 2 × 10−4
rad, undetectable at LIGO/Virgo/KAGRA but accessible
at Einstein Telescope / Cosmic Explorer.
Falsifier: ET/CE BNS waveform measurement of subleading soft-graviton coefficient outside [α2 · 0.9, α2 · 1.1]
window falsifies Theorem AH.203.
d.

CKM ρ̄ Path A Canonical Closure

Proposition AH.204 (CKM apex prediction
ρ̄ = 21.5α = 0.157 as candidate selection-rule extension).
The CKM apex prediction admits two structurally motivated assignments: Path A with ρ̄ = 21.5α = 0.157 from a
Cayley +5/2 correction, and Path B with ρ̄ = 22α = 0.161
from integer reassignment of the bundle filtration. Against
PDG 2024 (ρ̄ = 0.1591 ± 0.0094) the two paths are
statistically indistinguishable (Path A at 0.23σ, Path B
at 0.15σ), and neither is uniquely forced from the master
derivation tree at theorem grade; the Path-A preference
rests on the structural conditions discussed in the proof,
not on the data. Both paths are recorded as candidate
selection-rule extensions, conjectural rather than derived,
pending decisive measurement at LHCb Run 4 + Belle II
by 2030.
Plausibility argument. The anomaly-cancellation framework offers a candidate apex-correction formula of the

62
P
form δapex = ( i di − 1)/2 summing over McKay-quiver
representation dimensions. Two specific assignments yield
half-integer corrections compatible with the observed window: an apex-cycle assignment summing to 6 gives δ =
5/2 (Path A), while an alternative bundle-filtration assignment yields the integer 22 (Path B). The selection between
assignments is not currently forced by the master derivation tree. The group A5 has irreducible-representation dimensions {1, 3, 3, 4, 5} summing per Burnside to |A5 | = 60,
with no 2-dimensional irrep; the binary icosahedral cover
2I = SL(2, F5 ) has irreps {1, 2, 2, 3, 3, 4, 4, 5, 6} with
two distinct 2-dimensional irreps and two distinct 3dimensional irreps, so any apex triple drawn from 2I
involves a choice among multiple candidates of the same dimension. Tightening either assignment to theorem grade
requires a structural forcing argument that is not in hand;
both Path A and Path B are therefore kept as conjectural candidate selection-rule extensions, as is the original
Cayley-graph closure (status remark in Sec. AH 7 a). We
mark Path A as the preferred candidate on the basis of
(i) the apex-triple combinatorics and (ii) compatibility
with the uniform consecutive-triple step-size r = 1 in the
master derivation tree (Theorem AH.188) up to the apex
correction; but “preferred” here is a presentation choice,
not a structural forcing.
Falsifier: LHCb Run 4 + Belle II 2030 measurement of
ρ̄ outside [0.145, 0.169] at > 3σ falsifies Path A. Outside
[0.149, 0.173] at > 3σ falsifies Path B. The two paths differ
by 0.0036, so separating them at 3σ needs σ(ρ̄) ≤ 0.0012
(Remark AH.143).

e.

Frame-Theorem ksplit Smooth Interpolation

Theorem AH.205 (Padé [1/1). smooth interpolation
between FRW and galactic-EFE regimes] The smooth
interpolation between aFRW
= 0 (cosmological regime,
ext
k < ksplit ) and the galactic-EFE regime (k > ksplit ) is
given by the Padé [1/1] in x2 :
fDFD (k/ksplit ) =

(k/ksplit )2
,
1 + (k/ksplit )2

with ksplit = 0.040 h Mpc−1 , a modeling choice (Assumption AH.80), taken as the geometric mean of kH (Hubblelength wavenumber, 0.0021 h Mpc−1 ) and kopt (opticalmetric correlation wavenumber, 0.78 h Mpc−1 ). The interpolation preserves: (i) energy-momentum conservation;
(ii) gauge invariance; (iii) reduces to ΛCDM exactly at
k = 0; (iv) reduces to deep-MOND galactic dynamics at
k → ∞.
Proof. The Padé [1/1] is uniquely selected by four constraints: Padé hierarchy consistency, correct ∼ x2 lowk limit, correct ∼ 1 − x−2 high-k limit, and slope 1/2
at the boundary. The action-level interpolation modifies only the response operator at the matching scale;
energy-momentum conservation, gauge invariance, and

the Bianchi identity are preserved at second order. Under the adopted Rest-Mass Channel the interpolation
does not enter linear growth (the remark after Assumption AH.80).

16.

Fortifying Theorems for Multi-Route
Derivations

This subsection presents ten fortifying results on the
closed form for α−1 , the structural derivations of Λ, the
bundle decomposition, the priming family, the Branch B
exponents, the gauge-bundle structure, the BMS asymptotic group, and the Completeness Theorem categorization. The framing of the multi-route α presentations
(Theorem AH.189) is sharpened in light of the formalized
independence theorem (Theorem AH.216), and the Keff
accounting is sharpened in Theorem AH.207 to distinguish
continuous-parameter freedom from discrete axiomatic
choice.
a.

Five Presentations of α−1

Theorem AH.206 (Five presentations of α−1 ). The
closed-form value of α−1 is presented by multiple structurally distinct mathematical engines: (i) the Connes
spectral triple, (ii) Berezin–Toeplitz quantization, (iii) the
optical-metric variational principle, (iv) anomaly cancellation, and (v) the Verlinde-modular U (1)60 Chern–Simons
Hopf invariant on the Hopf fibration of CP 2 × S 3 . Routes
(i)–(iv) belong to the heat-kernel / spectral family and
are connected by named functorial bridges (Bordemann–
Meinrenken–Schlichenmaier; Connes–Moscovici); route
(v) lies in a structurally distinct categorical family.
Proof. The independence relations between pairs of
routes (related-but-not-equivalent for routes within the
spectral/heat-kernel family; structurally distinct for
route (v)) are formalized in Theorem AH.216 under
the standard categorical definition of independence
(no full+faithful+surjective equivalence functor between
routes). The Bordemann–Meinrenken–Schlichenmaier
and Connes–Moscovici functorial bridges establish that
routes (i)–(iv) share substantial heat-kernel structure
while remaining distinct presentations; the categorically
distinct fifth route is the Verlinde-modular derivation,
route (v).
b.

Keff : Zero in the Microsector

Theorem AH.207 (Zero continuous microsector knobs;
four cosmological calibrations). Under the standard
physics definition of a free continuous parameter, the
micro
dimensionless DFD microsector tree has Keff
= 0. The
full physical corpus has four cosmological calibrations: the

63
measured vacuum coefficient R behind (H0 tP )2 = Cα57 ,
the amplitude As , the coefficient of ωb h2 , and the shared
astrophysical reionization depth τ ; the particle sector adds
the fitted Higgs prefactor, the imported factor ηq and the
Higgs quartic, conjectured to be 1/8 at a matching scale
that is not derived. The framework’s adopted axioms and
discrete choices are listed in the inputs tables of the main
volume (App. FK). The continuous measured coefficients,
the dimensionless microsector closure, and the discrete
choices must be counted separately in any Bayesian comparison.
Proof. A continuous parameter has Lebesgue prior measure; a discrete choice has counting-measure prior. The
dimensionless microsector contains none of the former, but
the normalization ledger contains the continuous inputs
listed above. Consequently the description-length/Bayesfactor numbers that price DFD as having zero continuous
parameters apply only to the microsector comparison and
may not be advertised as a full-corpus Bayes factor until the prior and likelihood for all of them are included.
The discrete-choice accounting remains separate and unchanged.
c.

Completeness Theorem Cardinality 34/1/5

Theorem AH.208 (Completeness cardinality). After
applying the partial-grade closure of Theorem AH.203,
accounting for the CKM apex at partial grade, moving
the failed optical-screen reconciliation (excluded by Theorem AH.201) to the open gaps as the underived size of the
two-clock ratio, and moving the charged-fermion residuals, which Theorem AH.200 does not close, to the open
gaps, the cardinality is X = 34 answered + P = 1 partial
+ K = 5 open gaps = 40. The single partial-grade entry
is the CKM apex prediction; the added open items are
the underived size of the two-clock ratio and the chargedfermion residuals. The tally is bookkeeping of a question
list; the inputs themselves are listed in the main volume
(App. FK).
Proof. The size of the two-clock ratio (Q9 ) is not derived
and the charged-fermion residuals are not closed, so both
sit on the open ledger, which gives 34/1/5 without changing the total.
d.

Bundle (a, n) = (9, 5) from χ = 60: a Consistency Check

Theorem AH.209 (Unique bundle decomposition
from HRR + SM-content). Given the HRR constraint
χ(CP 2 , O(a) ⊕ O⊕n ) = 60 and the SM 5-species matter content per generation, the bundle decomposition
E = O(9) ⊕ O⊕5 is uniquely determined. Keff contribution from bundle decomposition: zero.
Proof. HRR formula: χ(O(a) ⊕ O⊕n ) = (a + 1)(a + 2)/2 +
n = 60. SM 5-species constraint forces n = 5. Substituting: (a + 1)(a + 2)/2 = 55 ⇒ a2 + 3a − 108 = 0 ⇒ a = 9.

The minimal-padding axiom is a redundant restatement
of the SM-content forcing once the HRR constraint is
fixed. Keff = 0 confirmed. The theorem runs the count
from χ = 60 back to (a, n), so it is a consistency check,
not a derivation of kmax : the forward count, in which
a = Ngen q1 = 9 and n = 5 give χ = 60, is the conditional
determinant-line lemma of the main volume (App. F),
with q1 = 3 from the electroweak identification.
e.

Priming Family

Theorem
AH.210
(Priming
family
{30/31, 45/44, 57/58, 14/13}
from
BT
axioms
+ chirality). The four-element priming family
{Fe = 30/31, Fu = 45/44, Fd = 57/58, Fν = 14/13} (Theorem AH.182) is obtained from Berezin–Toeplitz axioms
plus chirality projection of the master Dirac operator
on CP 2 × S 3 . The sector-specific Toeplitz dimensions
de = 31, du = 45, dd = 58, dν = 14 are assigned by the
rules of the proof; the direction rule (multiplicative for
direct-Higgs, divisive for conjugate-Higgs) is read from
chirality. Keff contribution: zero.
Proof. de = kmax /2 + 1 = 31 from chirality-projected
half-spectrum + null channel; du = a · n = 9 · 5 = 45 from
bundle product (conjugate-Higgs sector); dd = kmax −
Ngen + 1 = 58 from generation-projected level cap; the
neutrino entry is dν = dim H 0 (CP 1 , O(a+4)) = a+5 = 14
from the Dirac overlap (main volume, App. X). Direction
rule from chirality projection: direct-Higgs sectors yield
(d − 1)/d, conjugate-Higgs sectors yield d/(d − 1), so the
conjugate-Higgs rule gives Fν = dν /(dν − 1) = 14/13, and
13 = dν − 1 is the primed channel count.
Remark AH.211 (The dν bookkeeping). The neutrino
entry is dν = 14, as App. X derives it, dν =
dim H 0 (CP 1 , O(a + 4)) = a + 5, with Fν = dν /(dν − 1) =
14/13 under the conjugate-Higgs direction rule; the
count 13 that also appears is the primed (null-channelremoved) count dν − 1. No printed result fixes the total
31 + 45 + 58 + dν , so that sum is not a cross-check. JUNO
separates 14/13 from the alternative 13/12, which moves
∆m231 by 1.2% (2.510 against 2.540 × 10−3 eV2 ), at four
standard deviations or more.
f.

Branch B Exponents: A Forcing Argument

Theorem AH.212 (A forcing argument for the Branch
B exponents 3/11 and 7/20). The Branch B neutrino
exponents k = α−3/11 and r = α−7/20 (Theorem AH.168)
are argued to follow from (an argument not yet verified;
the branch is chosen by the oscillation data, main volume,
App. FK): (i) integer 11 = dim H 0 (CP 1 , O(10)) from
Bott–Borel–Weil applied to chirality-projected CP 1 halfspectrum; (ii) integer 20 = kmax /3 = 60/3 Sugawara
double-cover sector count via Z2 × S3 orbit counting on
SU(2)58 simple objects; (iii) integer 6 = chiral generation

64
count ×2; (iv) integer 7 = chirality-gauged fermion-charge
sum per generation. The combinatorial identity 6/11 +
7/10 = 137/110 would follow from the same (3,2,1) bundle
origin.
Proof. Bott–Borel–Weil on chiral half-spectrum:
dim H 0 (CP 1 , O(k)) = k + 1; at k = 10 (uniquely fixed by
Branch-B locking conditions: Padé denominator degree,
Sugawara level threshold, priming closure), dim = 11.
Z2 ×S3 orbit counting on the 59 simple objects of SU(2)58
MTC at Sugawara level 58 = 60 - 2 yields 20 sectors.
Generation count ×2 from anomaly-doubling under
Z2 chirality. Charge sum 7 from Atiyah–Singer index
plus anomaly cancellation via the anomaly-cancellation
derivation.
g.

Five-Presentation Λ ∝ α57 MP4 Convergence

Theorem AH.213 (Five categorical presentations of the
cosmological-constant exponent 57). The exponent 57
in Λ ∝ α57 MP4 admits five categorical presentations, all
converging: (i) Lemma O.5 + topology (57 = Ω − Ngen
= 60 - 3); (ii) Wheeler–DeWitt eigenvalue from primeddeterminant ratio (asserted, not derived; §AH 12 b status note); (iii) photon-sphere CFT anomaly with c =
174/60 = 2.9; (iv) the spectral action’s vacuum (a0 ) term
read as α57 -suppressed, a conjecture (App. WS); (v) holographic Bekenstein bound saturation at Hubble horizon.
These are presentations, not five mutually independent
derivations: as the main volume and Theorems AH.216
and AH.218 establish for the parallel α-route case, the αpower routes share substantial heat-kernel structure and
reduce to three engines, at most two of them mutually
independent, not five. Keff contribution: zero.
Proof. Each route fixes the exponent 57 as the derived
consequence of (Ω, Ngen ) = (60, 3); the presentations are
not independent measurements of an overall coefficient.
The current dictionary is
ρΛ
3
(H0 tP )2 = Cα57 ,
= ΩΛ Cα57 ,
ρP
8π
where C = 0.87784 ± 0.00266 is measured in the acoustic
sector and ΩΛ is an observed input. The SH0ES-side
C = 1.027 ± 0.029, SH0ES’s 2022 H0 re-expressed as
(73.04/72.09)2 , is the local arm, 0.9σ from the C = 1 that
bound-matter clocks read under the two-clock postulate of
the main volume (App. ZS); it is not a derived coefficient.
Thus the theorem-grade content is the exponent and its
five categorical presentations; neither C nor ΩΛ is silently
upgraded to a topological derivation.
h.

(3,2,1) Gauge Bundle: Selected, Not Forced

Theorem AH.214 (SU(3) × SU(2) × U(1) from CP2 ×S3
+ bundle, selected). The Standard Model gauge group
SU(3) × SU(2) × U(1)Y is selected, not forced (Remark AH.215): Connes’ inner-automorphism theorem

applied to the DFD spectral data over CP 2 × S 3 with
bundle E = O(9) ⊕ O⊕5 gives End(E) = U (1) × U (5),
and the reduction to the Standard Model group imports
the Standard Model; the electroweak factor is the isotropy
group of CP 2 under one stated identification and colour is
an input (main volume, App. F). Tr(Y 2 ) = 10 over three
generations follows from the Standard Model hypercharges,
which are inputs (Remark AH.215). Alternative gauge
groups (SU(5), SO(10), E6 , E7 , E8 , PSU(3), Pati–Salam,
Trinification) are individually ruled out via Spinc obstruction, χtop obstruction, or chirality-grading obstruction.
Proof. Connes inner-automorphism: Aut(A)/Inn(A) =
End(E) modulo center. With E = O(9)⊕O⊕5 , End(E) =
U (1) × U (5). The chirality grading plus det = 1 reduction yields SU(5) × U(1)Y , which under the Standard
Model 5-species reduction gives SU(3)C ×SU(2)L ×U(1)Y .
Anomaly cancellation (U(1)3Y , U(1)Y · SU(2)2 , U(1)Y ·
SU(3)2 , gravitational, Witten) all close on the (3,2,1) bundle; SU(5) GUT fails Spinc on CP 2 ; SO(10) fails 16-spinor
compatibility with χtop = 3; etc. Tr(Y 2 )|1gen = 10/3 via
direct computation; three generations: 10.
Remark AH.215 (Two open defects in the proof above).
(i) Tr(Y 2 ) = 10. Direct computation with Y = Q − T3
and YQ = 1/6 (Slansky normalization) gives
X
d3 d2 Y 2 = 61 + 43 + 13 + 21 + 1 = 10
3 ,
1 gen

three generations: 10,
confirmed independently by the closed form
Tr(Y 2 )|1 gen = 11d/18 + 3/2 at colour d = 3.
(ii) Two defects in the proof above remain open.
First, “which under the Standard Model 5-species reduction gives SU(3)C ×SU(2)L ×U(1)Y ” imports the Standard Model at precisely the step where the gauge group
is meant to emerge; as written the reduction is assumed,
not derived. Second, End(O(9) ⊕ O⊕5 ) = U (1) × U (5)
has dim = 26, giving 25 after unimodularity, not 12;
and the intermediate SU(5) is a grand-unified group
that the isotropy-Schur argument rules out geometrically.
The gauge-group claim of Theorem AH.214 is therefore
recorded as selected, not forced. The electroweak part
follows separately, as the isotropy group U (2) of CP 2
under one stated identification (main volume, App. F, the
remark on the electroweak group), and colour and the
family content are inputs, as in the Standard Model.

i.

Independence Theorem for the α Presentations

Theorem AH.216 (Standard mathematical definition of
independence applied to the α presentations). Under the
standard categorical definition (two derivations are equivalent iff a full+faithful+surjective functor exists between
them; related iff a faithful or full functor exists but not
both; independent otherwise), no pair of the five presentations of α−1 in DFD is equivalent. The complete pairwise

65
classification — six related-but-not-equivalent pairs, four
independent pairs, a maximum mutually-independent subset of cardinality two, and an engine count of three — is
established by the sharper functor-level analysis of Theorem AH.218, which controls whenever the counts differ.
Proof. Pairwise analysis of the five α−1 presentations
with explicit functors and obstructions shows that no
equivalence functor exists for any of the ten pairs; the
route-by-route bridge structure, and the resulting pair
and subset counts, are computed in Theorem AH.218.
(Pairwise independence among several specific pairs does
not by itself yield a mutually-independent subset of that
size; see the framing note following Theorem AH.218.)
Under any stricter definition that disallows any functor
whatsoever, no two theorems within a working axiomatic
system would ever be independent; we therefore use the
standard categorical definition throughout.
j.

BMS Strict Identity at All Orders

Theorem AH.217 (DFD’s asymptotic-symmetry group
is identical to GR’s BMS at all orders). The asymptoticsymmetry group of DFD at scri+ is the BMS group, identical to asymptotically flat GR’s at all orders, not merely
isomorphic at leading order. The supertranslation algebra
coincides at every order in the 1/r expansion via inductive verification of the bracket relations. The BT cutoff
at kmax = 60 is an UV cutoff that does not affect external
soft physics or asymptotic charges.
Proof. Asymptotic Metric-Coincidence Lemma: g̃ = e2ψ η
shares the same conformal completion as η (same scri+ ,
same universal Carrollian structure) since ψ → 0 as
r → ∞ at O(1/r). The bms4 bracket relations (Lorentz,
mixed weight −1/2, abelian supertranslations) hold at every order via inductive argument: assume identity at order
1/rn , verify at order 1/rn+1 . The ψ-rescaling subgroup
that vanishes at scri+ does not produce new asymptotic
symmetries; the residual asymptotic-symmetry group
equals GR’s BMS exactly. Soft-graviton theorems (Weinberg, Cachazo–Strominger, Sen) all reproduced as BMS
Ward identities.
k.

Structural-Derivation Note

Note on structural derivation. Each of the 34
continuous Standard-Model and cosmological parameters
predicted in this document has a derivation chain from
the master action, the topological invariant Ω = 60 and
the anchor MP , together with the inputs of App. FK of
the main volume. External verification of the derivation
chains by independent reviewers is invited and is required
for a substantive non-numerological status to be established; the present document offers the citation chains
and intermediate derivations as the basis for that external
verification.

17.

Sharpenings and New Predictions

This subsection presents results, each with its grade:
sharpened forms of results stated elsewhere, readings of
three open-gap questions, and two new sharp empirical
predictions.
a.

Categorical Analysis of the α Presentations

Theorem AH.218 (Three structurally distinct mathematical engines, five categorical presentations, six functorial bridges). The five α−1 presentations documented in
Theorems AH.189 and AH.206 are organized categorically
as three structurally distinct mathematical engines, five
presentations, and six named functorial bridges:
• Engine A (Spectral / NCG): comprises Routes
1 (Connes spectral triple), 3 (optical-metric variational principle), and 4 (anomaly cancellation),
fused via the Chamseddine–Connes spectral action
and the Connes–Moscovici local-index formula.
• Engine B (Berezin–Toeplitz): Route 2, bridged
to Engine A via the Bordemann–Meinrenken–
Schlichenmaier theorem but not equivalent at finite
level.
• Engine C (Verlinde / MTC): Route 5 (Verlindemodular Hopf invariant), structurally distinct from
Engines A and B.
Among the ten pairs {(Ri , Rj )}, six are related-but-notequivalent via named functorial bridges and four are pairwise independent (no functorial bridge in either direction).
Pairwise independence among 4 specific pairs is distinct
from the existence of a 4-element mutually-independent
subset. The maximum mutually-independent subset of
{R1 , R2 , R3 , R4 , R5 } has cardinality two; the engine count
is three.
Proof. For each pair, the kernel and image of the bridging
functor (or its absence) is computed explicitly. Routes 1
and 3 are connected via the Chamseddine–Connes spectral action of which the optical-metric Maxwell action is
the leading a4 heat-kernel coefficient; the bridge is full but
has non-trivial kernel (the gravitational sector). Routes
1 and 4 are connected via the Connes–Moscovici localindex theorem; faithful but not full. Routes 1 and 2 are
connected via Bordemann–Meinrenken–Schlichenmaier
convergence Tk → Dspectral at k → ∞; faithful at finite k,
full only in the limit. Routes 2 and 5 share the number
60, route 2 as its truncation kmax and route 5 as the
shifted level kWZW + 2 of the Verlinde formula at the
Sugawara level kWZW = 58; bridged but not equivalent
(different polarization classes). Routes 3, 4 and Route
5 admit no canonical equivalence functor and are independent. The maximum mutually-independent subset of
{R1 , R2 , R3 , R4 , R5 } has cardinality two (not five); the
engine count is three.

66
Scope. The categorical analysis fixes the structure of
the five presentations: the routes are not five mutually
independent derivations but three structurally distinct
mathematical engines presented in five forms with six
functorial bridges.

parameter Gaussian-vs-narrow-PDG priors, ∆K rises to
∼ 2360 bits. The width of the range reflects prior-volume
sensitivity, not a sharp determination of the comparison.

c.
b.

Solomonoff–MDL Bayesian Model Comparison

Theorem AH.219 (Solomonoff/MDL description-length
comparison). Under the Solomonoff/MDL Bayesian
framework, the description-length difference between
DFD’s dimensionless microsector and the Standard
Model’s continuous parameters, a microsector comparison that leaves out the particle-sector inputs of Theorem AH.207, satisfies
∆K ≡ K(SM) − K(DFD) ≈ 770 bits,
where K(·) denotes Kolmogorov / minimum-descriptionlength complexity. With K(DFD) ≈ 22 bits (axiom statement plus ∼ 4–7 discrete axiomatic choices each contributing log2 (Ni ) ≈ 2–4 bits) and K(SM) ≈ 790 bits (26
continuous parameters at ∼ 30 bits each plus structural
overhead), the conservative central estimate is
∆Kconservative ≈ 770 bits.
The range across reasonable prior choices is ∆K ∈
[660, 2360] bits, with the wide upper bound reflecting the
dominant role of prior-volume choice in MDL-based model
comparison rather than a sharp determination of the comparison.
Proof. A
continuous
free
parameter
under
standard
physics
priors
contributes
log2 (prior volume/measurement precision) ∼ 30 bits
to the descriptive-length penalty (Cover–Thomas
§14.3). A discrete axiomatic choice from a finite set
of cardinality N contributes log2 (N ) ∼ 2–4 bits per
choice under counting prior. The two categories are
mathematically distinct (Lebesgue vs. counting measure)
and information-theoretically distinct.
For DFD, the discrete-axiomatic choices are: (i) compact 7-manifold selection (effectively N ≈ 16 admissible candidates after Spinc + χtop = 3 + product structure constraints; ∼ 4 bits); (ii) bundle minimal-padding
(a, n) = (9, 5) (forced by HRR + SM 5-species, 0 free
bits but ∼ 3 bits of axiom statement); (iii) gauge structure SU(3) × SU(2) × U(1) (selected, not forced: its
electroweak factor rests on one stated identification and
colour is an input, Theorem AH.214; ∼ 4 bits of axiom statement); (iv) dimensional anchor MP (∼ 1 bit).
Total Kchoices ≈ 7 bits; total axiom statement ≈ 15
bits; K(DFD) ≈ 22 bits. For the Standard Model,
26 × 30 ≈ 780 bits of continuous parameters plus ∼ 10
bits of structural overhead gives K(SM) ≈ 790.
∆K ≈ 770 bits at the conservative central choice. Sensitivity analysis under hostile priors (parameter ranges at
10−30 to 1030 , discrete-choice menus at N ∼ 106 ) gives
∆K ≳ 660 bits for the choices counted above; under per-

Burnside Identity Forced at Representation-Theoretic
Level

Theorem AH.220 (Uniqueness of A5 among order-60
groups satisfying
constraints). The Burnside idenP DFD
2
tity |G| =
i di at |G| = 60 admits exactly one
solution among finite groups satisfying DFD’s structural constraints (non-abelian, simple, exactly five irreducible representations): namely G = A5 with
irreducible-representation
dimensions {1, 3, 3, 4, 5} sumP
ming to i d2i = 1 + 9 + 9 + 16 + 25 = 60.
Proof.
P 2 Burnside’s theorem: for any finite group G, |G| =
i di over the irreducible complex representations of G.
There are 13 groups of order 60. Among these, only
A5 is non-abelian and simple (a Sylow-theorem elementcounting argument shows A5 embeds into S5 as the unique
order-60 simple subgroup). Among non-abelian simple
groups, A5 is uniquely characterized by its five-irrep decomposition {1, 3, 3, 4, 5}.
Sharp falsifier. Discovery of a fourth chiral-fermion
generation at any collider would break the HRR Eulercharacteristic computation. This is a structural falsifier
of the framework.

d.

Pre-Registration Tier Classification of Observational
Signals

Theorem AH.221 (Three-tier classification of DFD observational signals). The five observational signals compared with DFD’s values in 2021–2026 data are classified
into three tiers by pre-registration status:
• Class A (pre-registered): the structural prediction was recorded prior to the published measurement. This class contains: ATLAS Higgs selfcoupling −1.2 < κλ < 7.2 (DFD’s value, κλ = 1,
is the Standard-Model one at a matched quartic, a
consistency check rather than a discriminator).
• Class B (structurally derived, observedafter): the prediction traces to a first-principles
structural derivation in earlier supporting records
that predates the published measurement. This class
contains: NA62 K+ → π + ν ν̄, HERA Phase I 21cm, FLAG 2024 chiral condensate.
• Class C (post-hoc consistency check): the specific numerical agreement was identified after the
data was published. This class contains: NuFIT 6.0
mass splittings (the neutrino branch is chosen by
the oscillation data, main volume, App. FK).

67
Proof. For each signal, the audit traces the date and form
of the DFD prediction and compares to the date and
form of the published measurement. The four Class A
and Class B signals trace to first-principles derivations (in
this document or earlier supporting records) that predate
the relevant published measurements. For the Class C
signal, the numerical agreement was identified after the
data were published.
Implication. Each signal carries its tier label: “DFD
predicts” for Class A, “DFD is consistent with” for Class
B, “post-hoc consistency check” for Class C.

e.

The Origin Question and the Wheeler–DeWitt State

Theorem AH.222 (Wheeler–DeWitt cosmogenic
state and the origin question). The proposed cosmogenic Wheeler–DeWitt ground state Ψ0 [ψ, Ξ] =
N exp(−SE [ψ, Ξ]/ℏ) (Theorem AH.177; the closed-form
eigenfunction and eigenvalue steps are asserted, not derived, per its status note, §AH 12 b) is DFD’s reading of
the origin question; it rests on two stated inputs, the noboundary state and the smooth beginning (main volume,
App. Q), and the origin itself stays open.
Proof. Contrast with rival cosmologies: inflation requires
a pre-inflationary state and defers the origin question;
loop quantum cosmology has a bouncing prior phase
(which contains its own origin question); cyclic models
posit infinite regress; multiverse selection requires a metatheory selecting the realised universe. DFD’s answer rests
on the proposed cosmogenic state and two stated inputs,
the no-boundary state and the smooth beginning, which
carries the arrow of time (Theorem AH.185).
Empirical signatures. The tilt ns = 1 − 2/60 =
0.96667 is a stated number; the mechanism is open
(main volume, App. FK), and the combination of Planck,
ACT DR6, CMB lensing and DESI gives 0.974 ± 0.003,
about 2.4σ above it; As is one calibrated amplitude,
matched by 32πα5 ; no detectable primordial B-modes
(r ≲ 10−7 ; main volume, App. SM); the dictionary’s
equil
fN
L ≈ 0.007 has no printed source and is not a prediction.

f.

Anthropic-Principle Redundancy Theorem

Theorem AH.223 (Structural completeness renders the
anthropic principle explanatorily unnecessary). For every
continuous parameter ci in the DFD master derivation
tree (Theorem AH.188), there exists a derivation chain
from the master action, Ω = 60, MP and the stated inputs
of App. FK to ci , without invoking observer-dependent
selection. The anthropic principle is therefore explanatorily redundant for the derived exponents and ratios; the
continuous inputs of App. FK remain.

Proof. Direct enumeration of the 34-parameter master
derivation tree (Theorem AH.188). For each ci (α, Λ, v,
charged-fermion masses, neutrino masses, mixing angles,
gauge couplings, etc.), the citation chain to the master action is exhibited. The running factors of Theorem AH.200
are not Standard-Model running, so they do not reduce
the charged-fermion residuals, and the table rests on the
stated inputs of the main volume (App. FK). Anthropic
reasoning posits selection from a multiverse ensemble;
DFD provides explicit derivation, eliminating the need
for ensemble-plus-selection. The redundancy is not refutation: anthropic reasoning may remain coherent in other
frameworks, but it does no explanatory work in DFD for
the derived exponents and ratios.
Consequence. DFD does not refute the anthropic
principle; it makes the principle explanatorily unnecessary
for the parameters in its domain.

g.

String-Landscape Exclusion Theorem

Theorem AH.224 (DFD’s structure precludes the flux–
compactification mechanism producing the string-theory landscape). The DFD configuration space has
zero-dimensional vacuum moduli (Theorem AH.179,
dim V / ∼= 0, |V / ∼ | = 1) and zero continuous microsector knobs (Theorem AH.207). The measured clock
coefficient C is a scale-dictionary input, not a vacuum
modulus. Therefore the flux-compactification construction
that produces the ∼ 10500 string-theory landscape cannot
be realized in any model satisfying the DFD axioms.
Proof. The string-theory landscape arises from integer
flux quanta and continuous complex-structure moduli on
Calabi–Yau manifolds, integrated over to produce a highdimensional vacuum moduli space. DFD’s eight rigidity
layers (Theorem AH.179) forbid: (a) the Calabi–Yau internal manifold (replaced by CP 2 × S 3 , unique under
the seven axioms of the main volume’s Theorem AB.1);
(b) flux-quantum freedom (the bundle decomposition is
uniquely (a, n) = (9, 5), Theorem AH.209); (c) continuous freedom in the gauge group (the (3, 2, 1) group is a
discrete selection, not forced, Theorem AH.214, and carries no modulus); (d) continuous-moduli vacuum freedom
(zero by Theorem AH.207). The construction is therefore
precluded structurally; no hidden landscape can satisfy
the eight rigidity layers simultaneously.
Sharp falsifier. Demonstration of a two-parameter
family of vacua satisfying all DFD axioms (eight rigidity
layers) would refute Theorem AH.179 and reopen the
landscape question.

h.

QCD Axion Structurally Excluded

Theorem AH.225 (DFD admits no propagating QCD
axion). Three independent walls each forbid a propagat-

68
ing Peccei–Quinn axion in DFD: (i) the gauge group
SU(3) × SU(2) × U(1), selected rather than forced (Theorem AH.214), has no second U(1)PQ factor; (ii) Theorem AH.209 (bundle decomposition uniquely (a, n) =
(9, 5), no axion sector); (iii) the microsector Hilbert space
saturation at kmax = 60 (no spectral room for a Goldstonelike mode).
Proof. The selected gauge group carries no U(1)PQ factor;
the Connes reduction from End(E) = U (1) × U (5) to
it imports the Standard Model (Remark AH.215). The
bundle decomposition is uniquely determined by HRR
plus the SM 5-species constraint with no integer-6-species
solution. The Berezin–Toeplitz Hilbert space is finitedimensional with dim = 60; an axion would require a
60th-mode-violating direction.
Sharp falsifier. Any haloscope detection of an axion
(ADMX-G2, MADMAX, ABRACADABRA, CASPEr,
IAXO, DMRadio) anywhere in the (ma , gaγγ ) plane falsifies DFD with no patch.

i.

j.

Higher-Form Symmetries of the Microsector

Theorem AH.227 (Discrete higher-form symmetries
of the SU(2)58 microsector). The DFD microsector with
SU(2)58 modular tensor category admits a Z2 one-form
symmetry generated by the simple current J = a29 (the
unique simple current in the MTC), and an integer twoform symmetry Z generated by the second cohomology of
CP 2 .
Proof. The SU(2)k MTC has simple currents indexed
by Z2 at integer level k (the unique non-trivial simple
current is the spin-k/2 object); for k = 58, this is the
a = 29 object. The two-form symmetry follows from
H 2 (CP 2 ; Z) = Z.
Falsifiers. The higher-form symmetry constrains
DFD’s allowed BSM extensions: (F1) any process violating the Z2 one-form symmetry at > 3σ falsifies; (F2)
any process violating the Z two-form symmetry.

Muon-to-Electron Mass Ratio

Theorem AH.226 (Lepton mass ratio from priming family and Cayley-graph distance). The structural prediction
for the muon-to-electron mass ratio is
mµ
31 3 −1.0
· ·α
· ηµ/e ≈ 212.4 ηµ/e ,
=
me
30 2
where 31/30 comes from the lepton-sector priming factor (Theorem AH.210), 3/2 comes from the generationoperator ratio Aµ /Ae = 1/(2/3), α−1.0 comes from the
Cayley-graph integer distance ne − nµ = 1.0, and ηµ/e
is a finite QED correction. CODATA 2022 measures
mµ /me = 206.7682830 ± 0.0000046; matching it needs
ηµ/e = 0.973, the residual-removal factor that Theorem AH.200 shows running does not supply.
√
Proof. The master mass formula mℓ = Fℓ Aℓ αnℓ (v/ 2)
(Theorem AH.182) gives the ratio
mµ
Fµ Aµ α n µ
=
,
me
Fe Ae αne
√
in which MP , v/ 2, and the universal
QED running
√
between the matching scale Λtop = α MP and the leptonic mass scale all cancel exactly. The priming ratio
is Fe−1 = 31/30 (gen-1 lepton has priming, gen-2 lepton does not at this order). The generation-operator
ratio is Aµ /Ae = (2/3)−1 = 3/2 from the discrete
generation count. The Cayley-graph distance gives
ne − nµ = 1.0, hence αnµ −ne = α−1.0 . Combining:
mµ /me ≈ (31/30) · (3/2) · 137.036 ηµ/e ≈ 212.41 ηµ/e .

k.

The a6 Heat-Kernel Coefficient

Remark AH.228 (a6 on the joint trace, and the exact
construction). On the joint seven-dimensional trace of
CP 2 × S3 the a6 coefficient does not vanish. For the
U(1) twist of the closed form for α, a6 /a4 = 19/5 (14/5
from CP 2 , 1 from S 3 ), which at the printed cutoff is 25%
(Gaussian) to 38% (sharp) of the a4 term, and up to 59%
for smooth plateaus (computed). A sub-ppb evaluation of
a6 on this space does not hold: its pure-curvature block
does not vanish on symmetric spaces, the Gilkey–Branson
prefactor 1/(5040 (4π)7/2 ) is 2.8 × 10−8 , and |a8 /a6 | is
several percent. The closed form for α does not need a6 : it
is the exact construction of the main volume’s conventionlocked closed form, the ζ(0) of the CP 2 charge trace times
the terminating heat trace of the ρ-shifted Casimir on S 3
with the heat-kernel regulator, which has no term beyond
the ones it keeps and is exact under those two stated
choices.
Significance. The closed form for α (Theorem AH.167)
and the exponent of Λ ∝ α57 MP4 do not rest on an a6
evaluation.

l.

Geometric Langlands Correspondence on the DFD Slice

Theorem AH.229 (Geometric Langlands functor on the DFD slice, conjectural). The DFD slice admits a Geometric
Langlands functor




1
1
L : D-mod Bunpar
SU(2) (CP ; {0, ∞}) −→ QC LocSO(3) (CP \ {0, ∞})

69
at level k = 58 with quantum-deformation parameter q = exp(2πi/60). The level k = 58 is read from the Atiyah–
Singer index identity 60 = k + h∨ for SU(2) (h∨ = 2). The DFD spectral data (AF , HF , DF ) on the gauge bundle
E = O(9) ⊕ O⊕5 over CP 2 supplies a canonical σDFD ∈ LocSO(3) whose Hecke eigensheaves L−1 (σDFD ) are in bijective
correspondence with the 59 = k +1 simple objects of the modular tensor category Repq (SU (2))58 (the integer 60 = k +h∨
is the Sugawara level shift, not the simple-object count).
Proof. The DFD slice fixes the curve as CP 1 with two marked points {0, ∞} (the parabolic moduli space has real
dimension 4, fibered as a 3-generation fibre over a 1-dimensional Yukawa-overlap base). The relevant DG categories
b 2 at level 58. The
on both sides are constructed via the Beilinson–Drinfeld center construction with chiral algebra sl
functor L is constructed via the Frenkel–Gaitsgory W -algebra at level 58 together with the Feigin–Frenkel duality, then
specialized to q = exp(2πi/60). Quantum deformation q = e2πi/60 matches Uq (sl2 ) at level 58 by the Kazhdan–Lusztig


b 2 , identifying the DFD modular data with the SU(2)58 Reshetikhin–Turaev
equivalence Repq (SU (2))58 ∼
= Rep sl
58
MTC. Compatibility with the Connes spectral triple is established by the explicit Yukawa-matrix → σDFD dictionary
derived from the Connes–Moscovici local-index formula. The integer 60 is unified across two functorial roles: level
shift k + h∨ and Atiyah–Singer index. (The MTC simple-object count for SU (2)58 is k + 1 = 59, structurally distinct
from these.)

Status. The result is conjectural; the published frameworks below do not contain it. The proof relies on the
published Beilinson–Drinfeld / Frenkel–Gaitsgory framework plus the Arinkin–Gaitsgory 2024 full-equivalence
theorem; the DFD-specific contribution is the slice identification, the compatibility with the Connes spectral triple,
and the unification of 60 = Ω across the two roles in the
proof.

Status. Conditional. The proof relies on published
Atiyah–Hitchin–Singer 1978, LeBrun 1989/1991/1996,
Connes 1996/2013, Witten 1989/2003, and Costello
2018/2021 frameworks; the DFD-specific contribution is
the five-step explicit construction and the spectral-action
↔ holomorphic-Chern–Simons intertwiner at level 58.

n.
m.

The Continuum Limit via Bochner–Minlos

Twistor Encoding (Conditional)

Theorem AH.230 (LeBrun twistor space functor on the
DFD slice). The DFD spectral data (AF , HF , DF , JF , γF )
on CP 2 lifts via the LeBrun twistor functor to the twistor
space Z(CP 2 ) = SU(3)/T 2 (the complete-flag manifold
of CP 2 , a Fano 3-fold of index 2). The Connes–LeBrun
functor F1 : SpecDFD → TwDFD is functorial and intertwines the Chamseddine–Connes spectral action with the
level-58 holomorphic Chern–Simons functional.
Proof. The construction of F1 proceeds in five explicit
steps: (i) Connes 1996/2013 reconstruction yields the
unique commutative spectral triple over C∞ (CP 2 ) from
the spectral data (AF , HF , DF , JF , γF ); (ii) LeBrun 1996
functor lifts the conformal structure of CP 2 (anti-selfdual for the reversed orientation) to its twistor space
Z = SU(3)/T 2 ; (iii) Borel–Weil–Bott diagonal lift identifies the bundle E = O(9) ⊕ O⊕5 over CP 2 with a holomorphic structure on Z; (iv) Atiyah–Hitchin–Singer 1978
/ Penrose–Ward 1976 correspondence matches the Dirac–
Dolbeault spectrum on Z with the spectrum of DF on
HF ; (v) the spectral action Tr f (D/Λ∗ ) is intertwined
with the level-58 holomorphic Chern–Simons functional.
Functoriality follows from the universal property of the
LeBrun lift.
The functor F1 is one leg of the Atiyah–LeBrun–Witten
triality of Theorem AH.176; the present theorem states
that leg alone.

Theorem AH.231 (The DFD continuum measure µ∞ ,
as asserted). The family of regularized DFD measures
{µΛBT } on the DFD covariant configuration space, asserted at finite Berezin–Toeplitz cutoff ΛBT in Theorem AH.190, is tight in the sense of Prokhorov. The
family admits a weak-limit cylindrical probability measure µ∞ on the nuclear dual S ′ (MDFD ) as ΛBT → ∞.
The limit measure µ∞ satisfies the Osterwalder–Schrader
axioms (OS0)–(OS4) and is regularization-independent.
Proof. By Bochner–Minlos (Reed–Simon Vol. II Theorem
IX.1, Glimm–Jaffe Theorem 3.4.2), a positive-definite
continuous functional Φ on the Schwartz space S(MDFD )
corresponds to a unique cylindrical measure on the nuclear
dual S ′ . The finite-ΛBT characteristic functional ΦΛBT is
established positive-definite and continuous via the BT
spectral representation. Tightness of the family {µΛBT }
follows from a Markov-inequality argument using weight
(1 + |x|2 )−3 together with Rellich–Kondrachov compactness: compact balls in the Sobolev space H −NS −4 are
compact in S ′ in the strong-dual topology. By Prokhorov’s
theorem there exists a weakly convergent subsequence;
uniqueness of the characteristic functional limit forces
full-sequence weak convergence to a limit µ∞ with characteristic functional Φ∞ = limΛBT →∞ ΦΛBT .
Each Osterwalder–Schrader axiom is verified at finite
ΛBT via the Lehmann–Källén spectral form of the BTregularized propagator and passes to the limit by dominated convergence: (OS0) translation covariance; (OS1)

70
reflection positivity from the Lehmann spectral representation; (OS2) Euclidean invariance; (OS3) regularity
(continuous moments); (OS4) ergodicity / uniqueness of
the vacuum.
Regularization independence: by the Padé–Stieltjes
uniqueness theorem applied to the BT spectral expansion,
any two admissible BT-cutoff schemes α, β give charac(α)
(β)
teristic functionals related by Φ∞ = Φ∞ ; hence µ∞ is
independent of regularization choice.
Scope. The Osterwalder–Schrader properties are asserted at finite cutoff and carried to the limit by dominated
convergence; for the interacting gauge sector this would
be a constructed four-dimensional measure, which the
main volume does not claim (App. YM).

o.

Two-Loop Graviton-Scattering Coefficient

Theorem AH.232 (Two-loop coefficient c2L for graviton
2 → 2 scattering, as asserted). The two-loop coefficient
in the DFD graviton 2 → 2 scattering amplitude is
c2L = (2.95 ± 0.07) × 10−6

in κ6 units,

computed via Berezin–Toeplitz quadrature on Hmicro =
C60 . The corresponding cumulative phase shift on a
binary-neutron-star late inspiral is ∆ϕcumulative
≈ 5×10−8
2L
rad, below the projected phase sensitivity of Einstein Telescope and Cosmic Explorer (a forced but unobservably
small correction). Dimensional analysis is verified; the
Goroff–Sagnotti 209/2880 structure emerges with the BT
spectral sum ζBT (60) = 1.0142 ± 0.0008 replacing the
conventional 1/(D − 4) pole.

Proof. The tree amplitude at κ2 is the standard Berends–
Giele recursion result for graviton MHV. The one-loop
integrand uses the BCJ double-copy of the squared YM
amplitude with BT-regularized propagator GBT (P ) =
(i/P 2 ) χ60 (P ) where χ60 is the BT cutoff function. The
one-loop coefficient evaluates to c1L ≈ 0.503, approximately 0.6% above the GR baseline of 1/2 (the predicted
DFD shift).
At two loops, three planar and four non-planar topologies contribute. The Goroff–Sagnotti 209/2880 rational structure emerges from the divergent part of the
conventional GR R3 counterterm; in DFD the divergence is replaced by the finite Berezin–Toeplitz spectral
sum ζBT (60) = TrHmicro (DF2 )−1 . Direct computation
of the triple sum over SU(3)-symmetric irreducible representations on CP 2 × S 3 yields ζBT (60) = 1.0142 ±
0.0008, giving c2L = (209/2880) · ζBT (60) · κ6 /(4π)4 =
(2.95 ± 0.07) × 10−6 in κ6 units (note (4π)4 = 24937, so
(209/2880) · 1.0142/(4π)4 = 2.95 × 10−6 ).
Dimensional analysis: [c2L ] = κ6 as required. Crosscheck: in the limit kmax → ∞, ζBT (kmax ) → 1/(D − 4) as
expected, recovering the GR Goroff–Sagnotti divergence.
Phenomenological consequence: the 2-loop phase correction to a BNS late inspiral integrates to ∆ϕcumulative
≈
2L
5 × 10−8 rad over the last 0.1 s before merger, below the
projected phase thresholds of Einstein Telescope and Cosmic Explorer (forced, but not currently observable).

p.

Subleading Corrections to Trans-Planckian BH Formation

Theorem AH.233 (Three subleading orders of the trans-Planckian BH formation amplitude, conjectural). The
trans-Planckian black-hole formation amplitude in DFD admits the asymptotic expansion
 πs  h
 s −3/2

 2 2 
i

−2
1 57
s
1 MP ΛBT
|ABH (s)|2 =
exp
1
−
α
+
O(s
)
,
−
2
2
20
s
MP
MP2
MP2
with three explicit subleading corrections: a Solodukhin–Sen logarithmic finite-size term at O(MP2 /s) with coefficient
−3/2 fixed by 4D-gravity universality; a de-Sitter Boltzmann factor at O(α57 ), the order of the cosmological-constant
relation Λ ∝ α57 MP4 , with coefficient βdS = 1/2; and a [[60, 3, ≥ 4]] stabilizer-code projection term at O(1/NBH ) with
coefficient −1/20 = −3/60 (logical/physical qubits per code block).


Proof. The leading exponential exp πs/MP2 arises from saddle-point analysis of the BH-formation cross-section
(Banks–Fischler 1999). The first subleading correction at O(MP2 /s) is the Solodukhin 2011 finite-size logarithmic
correction with coefficient fixed by 4D-gravity universality; the BT cutoff at ΛBT keeps it finite. The second subleading
correction at O(α57 ) comes from the de-Sitter-like Boltzmann factor associated with the cosmological-constant vacuum;
the coefficient βdS = 1/2 is fixed by the requirement that the de-Sitter horizon temperature match TH = H0 /(2π) in
the asymptotic limit. The third subleading correction at O(1/NBH ) encodes the photon-sphere [[60, 3, ≥ 4]] stabilizer
code structure: −1/20 = −(logical qubits)/(physical qubits per code block). The Penington 2019 island formula is
consistent with all three corrections.

Empirical signature (conditional; does not apply under √
the adopted completion). The estimate
h3 /h+ ∼ 1/ 20 ≈ 0.22 for a third polarization mode in
BH echoes is a dimensional estimate, conditional on (i)

the echo train itself (absent in both bands: the remnant
has a Kerr horizon, argued beyond first order in the spin,
App. WS) and (ii) a far-zone scalar mode — which the
corpus’s tensor-only far-zone statement and the elliptic

71
constraint reading (App. WS, Axiom E) exclude. As written it stands in tension with the two-polarization falsifier
clause; under the adopted completion both conditions
fail (no far-zone scalar mode under Axiom E; no echo,
App. WS), so the estimate does not apply.

Appendix WS: Wave-Sector Closure: Elliptic Scalar,
Derived Gravitomagnetism, δrad = 0, and Ringdown

This appendix derives three elements of the wave sector: (i) the radiative status of the scalar sector, δrad = 0;
(ii) the gravitomagnetic sector g0i , beyond PPN-template
matching; (iii) the relation between DFD’s physical coordinates and the standard PPN gauge. Two named inputs
carry the result, Axiom WS.1 and Proposition WS.5.
It then states the ringdown and echo items: with the
scalar-longitudinal piece σ of the parent strain kept, the
remnant of a merger is a Kerr hole (argued beyond first
order in the spin), it rings as Kerr, and there are no echoes
(Subsec. WS 5).

1.

Axiom E (Elliptic Constraint Reading)

Axiom WS.1 (E: the scalar sector is an instantaneous
constraint). In the preferred foliation (absolute time t),
the scalar sector obeys the elliptic AQUAL constraint at
each instant,


16πG
∇· µ(|∇ψ|/a⋆ ) ∇ψ = − 4 S[matter; ψ](t), (WS1)
c
with no independent temporal kinetic term in the local
(non-cosmological) sector; through first post-Newtonian
order the source S is the matter source of Lemma WS.3.
The absence of a temporal kinetic term, which makes the
action differ from general relativity’s by the (K − K̄)2
term of Proposition WS.6, holds, once baryons have decoupled from the photons, on the whole of every slice
on which the µ-law acts somewhere; while no baryons
are bound anywhere, and at least until baryons decouple from the photons, the scalar sector carries instead
the Einstein–Hilbert term’s own temporal term, and the
axiom selects its slicing (Remark WS.7). The one-field
form holds for systems without χ; inside bound systems
that contain χ the main text’s adopted energy of the two
Newtonian fields governs. Wherever the corpus writes
a c-speed d’Alembertian for ψ, that form is the numerical hyperbolic-relaxation of Eq. (WS1), not a claim that
ψ radiates (the channel-scope remark of App. AT). The
temporal functional K(∆) of App. Q remains the cosmological sector: ∆ is defined against the FRW screen flow
and governs homogeneous-deviation (dust-branch) physics
only.
For static matter in the µ → 1 band the source is
the main text’s optical-source law to all orders: on the
static matter metric, with the divergence taken in the
unimodular spatial part (eh )ij , that law is the time-time
component of the trace-reversed Einstein equation (computed), and with the parent strain’s σ kept, σ obeys the
traceless part of the Einstein equation and the Hamiltonian constraint follows (Proposition WS.6), so static
stars obey the equations of general relativity. Beyond
first order the source of moving or rotating matter also
contains the energy of the gravitational field, among it the

72
gravitomagnetic term ∂i (g0j Fij ) with Fij = ∂i g0j − ∂j g0i ,
and terms in ∂t ψ (computed); with the parent strain’s σ
kept the source is that of the maximal-slicing lapse equation and the slicing is maximal (Proposition WS.6(b),(e)).
Loop-induced higher-derivative terms of the matter fields
enter the constraint only through the renormalized source
⟨ε̂ + 3P̂ ⟩ren of the main text’s optical-source theorem, as
a relative correction of order (k/M̄P )2 ; written as terms
in ψ itself, their R2 part would supply, wherever the
(K − K̄)2 term acts, the only time derivatives and give
Eq. (WS1) a growing root at a rate of order (M̄P k)1/2
(computed).
Remark WS.2 (Why the elliptic reading is forced). A
no-go is proved: if the scalar sector is hyperbolic, the
far-zone breathing flux of a compact binary is δrad =
0.193 for PSR B1913+16 (monopole 0.015, cross 0.010,
quadrupole+trace 0.168; pipeline validated against Peters–
Mathews to 5 digits), excluded at ∼100σ; and no referenceflow choice in the K(∆) sector repairs this without shifting
the 1PN periastron advance by δ ω̇/ω̇ ≃ +0.064 (direct orbit integration), which breaks the measured mass triangle
at ∼90σ. The elliptic reading is the only structure specified in the corpus that survives both tests. Instantaneity
in the preferred frame is thereby a prediction carrier : it
is exactly why the scalar channel is radiation-free, while
all radiative physics lives in the (hyperbolic, luminal) TT
sector. Laplace-aberration safety is supplied by Theorem WS.9 below: matching EIH dynamics through 1PN
is precisely the statement that no anomalous aberration
effects appear at current sensitivity.

2.

The verified 1PN source and the Standard-Gauge
Coincidence Lemma

Lemma
WS.3
(1PN source,
point-particle
form;
machine-verified).
Direct
variation
of
R p
Spp = −mc2 dt e−ψ − eψ v 2 /c2 gives the conventionfree 1PN source
X ma c2 h
v2 i
S =
1 − 21 ψa + 32 2a δ 3 (x − xa ), (WS2)
2
c
a
(ψa : external field at a; self-terms renormalized). The
fluid form is S = 12 ρc2 [1+Π/c2 +2v 2 /c2 +2U/c2 +3p/ρc2 ]
(sympy-verified); the two agree through the proper-volume
factor e−3ψ/2 .
Lemma WS.4 (Standard-Gauge Coincidence). Solving
Eq. (WS1) with source (WS2) at 1PN (µ → 1) gives
ψ = 2U/c2 + ψ2 /c4 with
ψ2 = 4Φ1 + 4Φ2 + 2Φ3 + 6Φ4
(no χ̈ term: the solve is instantaneous),

(WS3)

so that
2U
c2
4Φ1 + 4Φ2 + 2Φ3 + 6Φ4 − 2U 2
+
,
c4
(WS4)


2U
(WS5)
gij = e+ψ δij = 1 + 2 δij ,
c
which is exactly the standard-PPN-gauge 1PN metric of
general relativity at γ = β = 1 with all other parameters
zero (no A term, no Whitehead term). DFD’s physical coordinates — the diagonal optical dictionary
plus absolute time — coincide with the 1PN standard PPN gauge, with no transformation and no
residue.
g00 = −e−ψ = −1 +

Proof. Green-function solution of ∇2 ψ2 = −8πGρ[Π +
2v 2 + 2U + 3p/ρ] term by term against ∇2 Φk =
−4πG (densities). The χ̈ term of the hyperbolic treatment
arises only from iterating c−2 ∂t2 ψ; Axiom E removes it at
the root. The Whitehead potential cannot arise: the field
equation’s only nonlinearity is the local, multiplicative eψ
in the source, and a local source under the flat Green
function never generates the non-local double integral
ΦW (Whitehead Locality Lemma). β = 1 is exact because ψ2 contains no U 2 piece; −2U 2 comes solely from
e−ψ = 1 − ψ + ψ 2 /2 − · · · .

3.

Vector-Sector Completion: g0i derived, not
matched

The Einstein–Hilbert term that the TT sector carries:
Proposition WS.5 (Spectral Einstein–Hilbert reduction). Let F be the truncated microsector, with Hilbert
space of dimension N (N = 60) and internal Dirac
operator DF , and let D = DM ⊗ 1 + γ5 ⊗ DF on
M4 × F . For a plateau cutoff f , constant near the origin, with every internal mass inside the plateau
(with G
√
fixed by Eq. (WS7) the top internal mass α MP sits at
m2 /Λ2 = αN f2 /(3π +f0 TrDF2 /MP2 ), at most 0.046 f2 , so
f2 /u0 ≲ 21.5 suffices for a plateau on [0, u0 ]; for a smooth
cutoff with f2 ≃ 3f0 and u0 = 1 that gives f0 ≲ 7, and
the internal masses of the near-arithmetic lattice relax it
to f0 ≲ 8.6; computed),
 D2 
Tr f 2
Λ
Z
N f4 Λ4 − f2 Λ2 TrDF2 + 21 f0 TrDF4
√ 4
=
gd x
4π 2
Z
N f2 Λ2 − f0 TrDF2
√
−
g R d4 x
48π 2
Z


f0 N
√
+
g 11E4 − 18 Cµνρσ C µνρσ d4 x
2
5760π
+ Sgauge + Sscalar ,
(WS6)

73
R∞
R∞
with f4 = 0 uf du, f2 = 0 f du, f0 = f (0) and E4 the
Gauss–Bonnet density. No local term beyond a4 appears;
this is an asymptotic statement, with a remainder that
falls faster than any power of Λ for curvatures and momenta well below Λ. The a2 term is the Einstein–Hilbert
functional of the four-dimensional metric, with
1
N f2 Λ2 − f0 TrDF2
.
(WS7)
=
16πG
48π 2
The result rests on two premises. (G1) The Einstein–
Hilbert term acts on the scalar field only through the
rest-mass channel of App. GR, that is, on the uniform
background and on the rest-mass perturbations, where it
replaces the printed Newtonian term rather than adding
to it; the µ-law acts outside it. The printed scalar action
is therefore the µ-bent scalar part of the Einstein–Hilbert
term: one term, not the Einstein–Hilbert term plus a
second one. G1 is not independent: its one term is the
static-energy identity of Proposition WS.6 with one G,
and its partition is the Rest-Mass Channel. (G2) Gravity’s
spectral trace uses the four-dimensional Dirac operator
D above, while the closed form for α uses its own trace
on the internal space (the regularization rule of the main
text’s convention-locked α); these are two different traces.
2
Proof sketch. Since D2 = DM
⊗ 1 + 1 ⊗ DF2 , the trace is a
sum over the internal eigenvalues λ2i of four-dimensional
heat-kernel expansions with shifted moments; on the
plateau f2 (x) = f2 − f0 x, f4 (x) = f4 − xf2 + 12 f0 x2 and
f0 (x) = f0 exactly, and every higher coefficient carries a
derivative of f at the origin, which vanishes. The truncation is exact in the four-dimensional sense on M4 × F ;
on a continuous internal space of odd dimension it would
not be. The Lichnerowicz formula gives E = −R/4, so
tr(E + R/6) = −R/3 per internal mode, which fixes the
attractive sign of G (an operator with E = 0 would give
the opposite sign), and the a4 coefficient of the Dirac operator contains no R2 term. Equation (WS6) reproduces
the exact Dirac spectral sums on the round four-sphere
and on S 2 × S 2 times a 60-mode internal space, with a remainder that falls faster than any power of Λ (computed);
its normalization agrees with that of Chamseddine and
Connes, the factor between 48π 2 and their 24π 2 being
the moment convention, with their mass trace equal to
half of TrDF2 .

The massless zero mode hµν (x) is the four-dimensional
metric itself, so its quadratic part is the linearized
Einstein–Hilbert action used at 1PN below and its cubic
vertex carries the tensor memory (App. AT). In strong
fields, where µ → 1, the Einstein–Hilbert term governs
the whole metric: with the parent strain’s σ kept, matter,
light and the tensor field share it, and the static exterior
is the Schwarzschild metric in DFD’s coordinates. The
tensor and vector sectors, and the scalar sector at µ → 1
(Lemma WS.4), share one G. Its value is not predicted:
it fixes N f2 Λ2 − f0 TrDF2 in Planck units, one scale measurement, as in
√ general relativity, and the internal masses,
57 modes at α MP in the printed spectrum, shift 1/G

by about 4%. The Weyl-squared
term implies an extra
p
spin-2 pole at m2 = Λ 10f2 /3f0 ≃ 3.6 M̄P for N = 60
and f0 = 1, above the cutoff, a tachyonic ghost under the
continuation that gives G > 0; it moves no observable
((k/m2 )2 ∼ 10−81 in the LIGO band), and the action
stops at a4 only below the cutoff. The a4 gravitational
term does not enter the Friedmann law: the Weyl term
vanishes on the Friedmann background and the Gauss–
Bonnet term is a total derivative. The displayed a0 term
is Planckian: on the near-arithmetic lattice of the internal spectrum, m2n = (n − 3) αMP2 /57 (n = 4, . . . , 60), its
4
term −f2 Λ2 TrDF2 /4π 2 alone is −(0.532 + 0.012f
√0 ) M̄P ,
and −(1.045 + 0.046f0 ) M̄P4 with all 57 modes at α MP
(computed). The dark-energy density is the α57 microsector term of App. AU with its coefficient R measured, and
reading it as the α57 -suppressed a0 term is a conjecture,
which needs the whole DF -dependent vacuum suppressed
with it. On the truncated microsector there are no Kaluza–
Klein modes; on a continuous K the Lichnerowicz count
(b1 (CP 2 ) = b1 (S 3 ) = 0, both factors Einstein-rigid) leaves
no extra massless spin-2 mode.
The spectral action and its Einstein–Hilbert term are
the construction of Chamseddine and Connes (Phys. Rev.
Lett. 77, 4868 (1996); Commun. Math. Phys. 186, 731
(1997)), and the idea that gravity is induced by the quantum fields of matter is Sakharov’s (Dokl. Akad. Nauk
SSSR 177, 70 (1967); Sov. Phys. Dokl. 12, 1040 (1968)).
What DFD adds is this: its own 60-mode finite space, the
structure that fixes α, supplies the finite geometry of that
construction; the Friedmann law and the nucleosynthesis
check follow (the optical-time Friedmann law of the main
text); the scalar sector and the Einstein–Hilbert term
are one term (premise G1); and the clock constant is
independent of G (App. AP).
Proposition WS.6 (Einstein–Hilbert completion of the
parent strain). Let the physical metric carry the full parent
strain of the main text’s gravitational-wave section: lapse
N = e−ψ/2 , a shift, and spatial metric gij = eψ (eh )ij
with hij the traceless part of the parent strain (hTT
ij , the
vector piece and the scalar-longitudinal piece σ), so that
N 2 (det gij )1/3 = 1 (the lock); let its dynamics be the
Einstein–Hilbert term of Proposition WS.5, with no temporal kinetic term for ψ (Axiom WS.1), in DFD’s absolute
time, in the rest-mass sector, where µ → 1. Then, for
isolated systems:
(a) the lock leaves only the spatial diffeomorphisms that
preserve the arena’s volume form; they remove the
vector piece but not σ, so σ is physical, and its
equation, from the traceless part of the Einstein
equation, is not a gauge condition;
(b) every solution has vanishing mean curvature of
DFD’s slices, K = 0: the absolute time is the maximal slicing;
(c) general relativity’s Hamiltonian constraint holds,
and the solutions are exactly those of general rela-

74
tivity in maximal slicing, written in the coordinates
of the lock;
(d) static stars obey the Tolman–Oppenheimer–Volkoff
equations; the exterior of a static star is the
Schwarzschild metric, e−ψ = 1 − 2GM/(c2 R)
at areal radius R, with ψ = 2GM/(c2 r) −
4
2 3
3 (GM/c r) + · · · in the arena radius r; a stationary rotating exterior is Kerr’s at first order in the
spin, and a spinning black hole’s is Kerr’s; through
second post-Newtonian order the two-body dynamics is general relativity’s; and for compact bodies
inertial and gravitational mass agree;
(e) the scalar equation is the maximal-slicing lapse equation, which for static matter is the main text’s
optical-source
divergence of (eh )ij :
 −h ij
 law with the
4
∂i (e ) ∂j ψ = −(8πG/c )(ε + 3P ) eψ ;
and on the Friedmann background:
(f ) linear perturbations keep the mean curvature of
DFD’s slices uniform, δK = 0 for k = 0 wherever no phantom field changes: the absolute time
is the uniform-expansion slicing, no scalar mode
propagates, the curvature perturbation ζ of adiabatic modes is conserved outside the horizon, and
for the curvature constant k = 0 of the first integral
the linear equations are general relativity’s.
The equivalence with general relativity in (b)–(d) and
(f ) is proved or computed for static configurations, for
linear perturbations of flat space and of the Friedmann
background, through second post-Newtonian order, and at
first order in the velocities to all orders in compactness;
for fully nonlinear time-dependent fields, for stationary
rotation beyond first order in the spin, and for maximal
slicing during generic rotating collapse, it rests on the
argument given for (b).
Proof sketch. (a) On the lock a spatial diffeomorphism


ξ at fixed slicing changes ln N (det gij )1/6 by 13 ∂i ξ i , so
only ∂i ξ i = 0 survives: two functions, which remove the
transverse vector piece; setting σ = 0 would take a third.
At first post-Newtonian order σ does not enter gij , so
the isotropic dictionary of Theorem WS.8 is exact there.
(b) Without a temporal kinetic term for ψ the action
differs from general relativity’s by a term quadratic in
K; general relativity’s action is invariant under normal
deformations of the slices, and the variation of the remaining term along them gives Di (N 2 Di K) = 0 on every
solution, so N > 0 and K → 0 at infinity force K = 0
(for actions whose momentum trace is proportional to
K, the Hamiltonian form of this identity is that of Henneaux, Kleinschmidt and Lucena Gómez, Phys. Rev. D
81, 064002 (2010), eq. (4.1)). This is computed for static
fields, for linear perturbations, and at first order in the
velocities to all orders in compactness, where the momentum constraint alone forces K = 0; for general fields
the identity, and the existence of the lock’s coordinates

on general data, are argued, not computed. (c) Every
term of the action is spatially covariant; with the matter, momentum, traceless and ψ equations imposed, the
spatial diffeomorphism identity
gives the Hamiltonian√
constraint density as c(t) det g N 2 , and asymptotic flatness sets c = 0; the term quadratic in K vanishes on
solutions, so the remaining equations are general relativity’s, and conversely general relativity’s maximal solutions,
solve them. For static fields

√ written with
√ the lock,
ˆ − 1 |∂ψ|2 , ĝ = eh , so
N det g (3)R = det ĝ R̂ − 2∆ψ
ĝ
2
the Einstein–Hilbert scalar energy is the printed Newtonian term with the same coefficient c4 /32πG: premise
G1’s “one term” (computed). (d) The exterior map is
r3 = R3 − 3GM R2 /c2 + C ′ , with C ′ set by the interior;
the reference Γ = 2, K = 100 polytrope has maximum
mass 1.637 M⊙ , the same computed in DFD’s coordinates
and from the TOV equations, and a moving neutron
star’s inertial mass equals its gravitational mass to better
than 10−8 up to the maximum mass (computed). At
second post-Newtonian order σ is sourced by the combination that sources general relativity’s conformal mode,
and its exchange cancels, at orders G, G2 and G3 , the
terms that the lock would otherwise add to the twobody Lagrangian (computed). (e) With N = e−ψ/2 the
lapse equation D2 N = N [Kij K ij + 4πG(ρ + S)] reads
∂i [(e−h )ij ∂j ψ] = −eψ [8πG(ρ + S) + 2Kij K ij ] (c = 1,
S the trace of the spatial stress), whose static form is
the stated law. (f) On the Friedmann background the
Einstein–Hilbert term keeps its Friedmann law, so the
term quadratic in K is quadratic in K − K̄, with K̄(t) the
uniform mean curvature of the slice itself, and the identity
of (b) becomes Di (N 2 Di K) = N K̄˙ (K − K̄); linearized
in conformal time for k = 0 it reads (Q − ∇2 ) δK = 0,
with ∇2 the comoving Laplacian, Q = 3(H2 − H′ ) and
H the conformal Hubble rate; Q > 0 for ordinary matter, so δK = 0. The three scalar equations, the lapse
equation, the momentum constraint, in which the shift
enters through the shear of σ, and general relativity’s
traceless Einstein equation, which determines σ, are then
Einstein’s in the uniform-expansion slicing; the trace, the
shift and σ are fixed by the matter at each instant, and
the constant c of (c) is proportional to the curvature
constant k (computed).
Black holes follow. The exterior’s event horizon is
Schwarzschild’s, at the areal radius 2GM/c2 , and Kerr’s
for a spinning hole, argued beyond first order in the spin.
For an eternal static hole it is the main text’s optical
horizon, the surface where n = eψ → ∞, which the lock
places at the arena’s origin; a hole formed by collapse
settles on the late-time maximal slicing, where the horizon
is a flow horizon, the inflow of the slices equalling the
local speed of light, with n = 256/27 on it, and n diverges
only on the slices’ limiting surface, R = 3GM/(2c2 ), at
the arena’s origin (computed).
The Boltzmann-code computations of the main text,
the χ third peak among them, solve general relativity’s
linear perturbation equations; by (f) these are DFD’s

75
own equations, in its uniform-expansion slicing, for the
curvature constant k = 0, which Remark WS.7 derives
and which the main text’s fits take; while no baryons
are bound, where the (K − K̄)2 term is absent, they are
the Einstein–Hilbert term’s own equations in the same
slicing. The temporal functional K(∆) of App. Q does
not enter them: it acts on homogeneous deviations only
(Axiom WS.1). Wherever the µ-law acts the equivalence
does not extend to it: its term lies outside the Einstein–
Hilbert term (premise G1).R Writtenpon the physical metric, that term is (a20 /8πG) d3 x N det gij [W (X) − X],
with X = c4 g ij ∂i ln N ∂j ln N/a20 the squared proper acceleration of the observers at rest √
in the slices of absolute
time in units of the epoch’s a0 = 2 α cH, with H read on
bound-matter clocks (App. ZS); it keeps the spatial diffeomorphisms that the lock leaves, at Newtonian order it
is the flat form of Axiom WS.1, and where µ ̸= 1 the two
differ at relative order (1−µ)ψ. It is the acceleration term
of the preferred-foliation theory of Blanchet and Marsat
(Phys. Rev. D 84, 044056 (2011); arXiv:1107.5264), written on DFD’s fixed slices.
Remark WS.7 (The curvature constant is zero). While no
baryons are bound anywhere the µ-law acts nowhere, so
under Axiom WS.1 the action is the Einstein–Hilbert term
alone. Its identity under a change of slicing then reads
ċ = 23 c ∂i β i , with c the constant of Proposition WS.6(c),
proportional to the curvature constant k of the first integral. For c ̸= 0 the identity fixes the longitudinal shift,
so no slicing keeps Eq. (WS1) elliptic, and ψ and σ carry
two modes of their own per comoving
√ wavenumber κ, one
of them growing at the rate κ/ 3 per unit conformal
time whatever the size of k, the matter content or the era,
with no upper limit in κ (computed). So k = 0 is the only
value that keeps the axiom and a well-posed evolution
before any baryons are bound. At k = 0 the equations
keep a2 δK constant before the (K − K̄)2 term acts, and
the momentum of ψ stays continuous where it switches on,
which sets δK = 0: the slicing is the uniform-expansion
slicing, K = K̄, throughout. Switched on there, with the
first phantom growing from zero, the term changes neither
the action nor the momenta, and c stays zero (computed;
after the switch at linear order). A detected spatial or
dark curvature falsifies this.
Theorem WS.8 (Vector-Sector Completion). Under
Proposition WS.5, the h0i components obey the linearized
EH momentum constraint. In the gauge fixed physically by
the DFD dictionary — (a) gij isotropic, = e+ψ δij through
first post-Newtonian order, (b) g00 = −e−ψ through the
same order, with ψ the elliptic solution of Lemma WS.4
— the unique 1PN solution is

1
g0i = − 3 72 Vi + 12 Wi
(WS8)
c
— the standard-gauge value of general relativity. The farzone coefficient −4, the Lense–Thirring precession, and
the leading spin terms of App. AX follow as consequences.
Beyond linear order the Einstein–Hilbert term couples the

shift to ψ: at O(JM ) it gives the stationary exterior the
general-relativistic factor 1 − GM/c2 r in gtϕ (App. AX).
Proof sketch. By Proposition WS.5 the zero-mode dynamics is GR, and at 1PN only its linearization enters: the
h0i equation is the momentum constraint sourced by ρvi .
Conditions (a)+(b) are precisely Will’s standard-gauge
conditions: (a) fixes the spatial gauge function; (b) — the
statement that all c−4 scalar content of g00 sits inside
e−ψ[elliptic] , i.e. no χ̈ term — fixes the temporal gauge
function (Lemma WS.4). With both functions spent, the
momentum constraint has the unique solution (WS8);
uniqueness is the standard 1PN elliptic uniqueness with
Vi , Wi → 0 at infinity. In DFD the “gauge” functions are
not free: coordinates are physical (preferred foliation),
and the dictionary conditions (a)+(b) are what define
those coordinates; the theorem states that the physical
frame the scalar dictionary selects is the same frame
in which the vector constraint takes the standard-gauge
form.
Condition (b) also fixes the slicing. For gij = eψ δij
the mean curvature of the constant-t slices is K =
3
− 2c
∂t ψ + ∂i g0i at leading order, and the standard-gauge
3
value (WS8) has ∂i g0i = 2c
∂t ψ, so K = 0: the preferred
foliation is the zero-mean-curvature (maximal) slicing at
this order. The longitudinal part of g0i is then set by
the slicing, and its transverse part obeys the Einstein–
Hilbert equation. At linear order the longitudinal part
of the momentum constraint, ∂t (∇2 ψ + 8πGρ/c2 ) = 0,
follows from the Poisson equation, so the two statements
agree there. Conditions (a) and (b) hold through first
post-Newtonian order; exactly, the lock fixes the lapse,
N = e−ψ/2 , and det gij = e3ψ , and with the parent
strain’s σ kept the slicing is maximal for isolated systems
(Proposition WS.6(b)).
Theorem WS.9 (EIH equivalence; binary-pulsar triangle). Lemma WS.4 + Theorem WS.8 reproduce the
full Einstein–Infeld–Hoffmann 1PN two-body Lagrangian.
Hence the periastron advance, Einstein delay, Shapiro
parameters, and the (ω̇, γE , Ṗb ) mass-triangle consistency
of PSR B1913+16 and J0737−3039 are GR-identical.
(The +6.4% periastron excess of the hyperbolic/K-branch
readings is absent by construction.)
Remark WS.10 (Data lock: J0737−3039 forces this result twice over). The published double-pulsar timing
(Kramer et al. 2021, PRX 11, 041050) gives Ṗb /ṖbGR =
0.999963(63) — a 1.3 × 10−4 test of the radiative sector — and measures ω̇ at 8 × 10−7 fractional precision.
Both pillars of this appendix are therefore independently
data-forced by the same system: (i) radiative: any scalar
breathing residual ≳ 10−4 of the GR flux is excluded, so
the elliptic reading’s δrad = 0 exactly (Theorem WS.11) is
the only DFD reading that survives — the K(∆)-branch
estimate (δrad ∼ 10−3 ) would fail this bound by ∼8×;
(ii) conservative: at 8 × 10−7 on ω̇, nothing short of the
exact EIH equivalence of Theorem WS.9 suffices. Both
bounds are met, with no tuning available or needed.

76
4.

δrad = 0 with a mechanism; scalar memory;
falsifiers

Theorem WS.11 (Radiative budget). Under Axiom E
and Proposition WS.5: the scalar sector, being an instantaneous constraint, carries no radiative flux in any
multipole (monopole, dipole, quadrupole breathing all identically zero as radiation); the TT sector carries the GR
quadrupole; therefore
δrad = 0

exactly at leading order,

(WS9)

with corrections only from µ-function effects where
|∇ψ|/a⋆ is not asymptotically large (a consequence of
the mechanism above, not an assertion).
Remark WS.12 (Scope: energy flux, not remnant spectroscopy). Theorem WS.11 governs the energy-flux channel — inspiral phasing and orbital decay (Ṗb ), where it
is confirmed at 1.3 × 10−4 (Remark WS.10). It does not
by itself fix the quasinormal-mode spectrum of a merger
remnant, which is set by the remnant’s metric; with the
parent strain’s σ kept that metric is Kerr’s, argued beyond
first order in the spin (Subsec. WS 5).
Corollary WS.13 (Scalar memory closes). The open
case of the App. AT channel-scope remark (head-on/burst
scalar memory) closes: a constraint field has no far-zone
radiative degree of freedom, so scalar memory vanishes
for all source classes; total memory is the tensor (TT)
memory, which equals GR’s, linear (ordinary) plus nonlinear (Christodoulou), because the TT sector is the full
a2 Einstein–Hilbert term (Proposition WS.5; App. AT).
Remark WS.14 (Consistency panel). (i) GW170817: cT =
c untouched (TT hyperbolic, luminal); on the optical-cone
reading of the TT sector the constraint is satisfied identically (GWs and photons ride the same cone). (ii) LVK
polarizations: no scalar/vector far-zone modes — consistent with the corpus falsifier clause; a confirmed far-zone
breathing mode falsifies Axiom E outright. (iii) Preferredframe causality: instantaneous constraint in a preferred
foliation generates no closed causal loops (absolute time
orders all events). (iv) Energy balance: Ṗb is TT-only;
radiation reaction enters the matter EOM through the
TT interaction (Burke–Thorne analog in the TT channel),
with the scalar sector conservative. (v) Frame-dragging
phenomenology at leading order (LAGEOS, GP-B, the
leading QPO spin terms) rests on Eq. (WS8); with the
parent strain’s σ kept the strong-field shift is Kerr’s, so
the spin-ISCO and the shadow displacement of App. AX
are Kerr’s, and the e−2u closed form of the first-order
shift gives way at O(JM ) to general relativity’s 1 − 3u.

5.

Ringdown and echoes: the optical and tensor
channels

a. Ringdown. Where µ → 1, with the scalarlongitudinal piece σ of the parent strain kept, light, matter

and the tensor field share one metric, whose dynamics is
the Einstein–Hilbert term (Proposition WS.5): the remnant of a merger is a Kerr hole in DFD’s coordinates and
its quasinormal modes are Kerr’s, beyond first order in the
spin by the argument of Proposition WS.6. Against the
ˆ
published constraints — GW250114: δf
220 = 0.02 ± 0.02
(90% C.L.); GWTC-4.0 hierarchical: 0.00 ± 0.06; damping
ˆ 220 = −0.01+0.10 — DFD’s δf220 = δτ220 = 0 is consisδτ
−0.09
tent, and the ringdown does not distinguish DFD from
GR. The eikonal
Ωc = c/bcrit is Schwarzschild’s,
√ frequency
2
since bcrit = 3 3 GM/c : the optical and tensor channels
ring together.
b. Echoes. The remnant’s Kerr horizon, argued beyond first order in the spin, has no reflecting surface, so
there is no echo train in either band and no third
√ polarization; the echo-polarization claim h3 /h+ ≃ 1/ 20 ≈ 0.22
(App. AH) does not apply.
c. The flux channel. Theorem WS.11 and Remark WS.10 govern the energy-flux channel (Remark WS.12) and are unchanged.

6.

Consistency notes on other sections

1. PPN vector sector. The pure-V solution Ni =
−4Vi /c3 of the shift equation is not transverse; the
transverse-gauge solution is Ni⊥ = −2(Vi + Wi )/c3
(since Vi⊥ = (Vi + Wi )/2 by the continuity identity; Vi and Wi already contain G). The two differ
by a longitudinal gradient, so no physical quantity
changes; Theorem WS.8 gives the vector sector,
and the PPN section states the transverse-gauge
solution.
2. TT coupling normalization appears as 16πG/c4
at every print (§formalism, §gravitational waves,
App. AT). The canonical value (fixed by the
Sh Rnormalization c4 /64πG and the interaction
ij
4
4
+ 12 hTT
ij T ) is 16πG/c ; a c /32πG prefactor or
a − 12 coupling is inconsistent with it (App. GR,
remark on the TT action normalization, item (ii)).
3. Echo polarization (App. AH vs. the
gravitational-wave section). Under Axiom E the
far-zone scalar mode is absent, and with the parent
strain’s σ kept the h3 /h+ ≃ 0.22 echo-polarization
claim does not apply: there is no third polarization
and no gravitational-wave echo (Subsec. WS 5).
4. EH attribution. a2 carries the Einstein–Hilbert
term (gravitational-wave section, App. O, App. SM);
a4 carries the gauge kinetic term and α (App. O)
— two different operators, not a convention choice.
This appendix uses a2 for the Einstein–Hilbert term
throughout (Proposition WS.5).
5. Ringdown and echoes: the statements of
App. AT, App. AH, App. RM and the strong-field

77
section follow Subsec. WS 5: Kerr ringdowns, argued beyond first order in the spin, and no echoes.

7.

What remains open in this sector

(1) A full nonlinear well-posedness statement for the
coupled elliptic-constraint + hyperbolic-TT system (an

open program item of the corpus; constraint-hyperbolic
mixed systems are standard in NR as “elliptic-hyperbolic”
formulations).