Lean 4 + Mathlib formalization of shadow holography, the organizing idea of the Golden Physics Project. It is a library of kernel-checked theorems, with every open step named explicitly as it is built.
Public site: https://lean.goldenphysics.org (plain-language overview) · technical blueprint at https://lean.goldenphysics.org/blueprint/ · site map
Source text: Daniel Toupin, On the Nature of Nature — https://goldenphysics.org
Author: Daniel Toupin | ORCID: 0009-0003-7682-9579
Exploratory companions: the five numeric workbenches GPPDiscovery, GPPDiscovery2, GPPDiscovery3, GPPDiscovery4, GPPDiscovery5 (numerical exploration that feeds the formalization)
One involution appears in two places.
- Celestial holography. Scattering amplitudes are Mellin-transformed in energy. They
are labelled by conformal dimensions
Δ, with the principal series atRe Δ = 1. The shadow mapΔ ↦ 2 − Δpairs each dimension with its partner. - Arithmetic. The completed zeta function
ξ(s)has a functional equations ↦ 1 − s, and its critical line sits atRe s = ½.
Under Δ = 2s these are the same map. The version that matters is the antiholomorphic
reflection τ(s) = 1 − s̄. Its fixed set is exactly the critical line
(GppFE.critical_line_is_fixed_locus), and ξ ∘ τ = conj ∘ ξ.
The framework reads this holographically, at three nested levels. All three are Mellin or Pontryagin dualities, so each is a precise dictionary rather than an analogy:
| Boundary | Bulk |
|---|---|
multiplicative line ℝ₊, split at its fixed point x = 1 |
spectral sphere, split by the critical line (the equator) into two hemispheres |
inversion x ↦ 1/x |
reflection s ↦ 1 − s̄ |
| support on a half-line | analyticity on a hemisphere (Paley–Wiener) |
| the integers as Fourier modes of a circle (theta, Poisson) | the functional equation |
primes as a compact torus ∏ S¹, dual to ℚ₊^× |
the critical line embedded densely in that torus (Kronecker): a hologram inside a hologram |
The same equator appears in celestial scattering as the scale sector: boosts act on energy
by dilation. The framework's central hypothesis is that ξ(s) is a physical operator, and
this library builds, theorem by theorem, the exact dictionary that hypothesis stands on.
The goal is to map the full arithmetic ↔ celestial CFT correspondence, carrying the dictionary as far as it goes. Each transfer is either kernel-checked or located at a specific, named step, and both are recorded here.
The library holds about 1,200 Lean modules. Each major directory carries its own README with the map of that layer; this section is the index of them.
- The dictionary itself. Shadow/functional-equation identification, dilation-unitarity
bridges between principal series and completed zeta, Haar self-duality, and the zero
pairing
ρ ↦ 1 − ρ̄. Most of the tree is inCelestialHolography/andRiemannHypothesis/. - The Tate layer. Tate's thesis for ℚ, built end to end: the local functional
equation at the real place for both parities, the ideles of ℚ and Fujisaki's theorem,
the standard adelic character with
ℚ^⊥ = ℚ(GP-DUAL), the Haar normalization with covolume one, the self-duality of∏_p ℤ_p, and the classical Poisson control made unconditional. The guide to the layer, with every file, its principal results, and its dependencies, isdocs/TATE-GUIDE.md. Fourier inversion and Plancherel on the adelic Schwartz class are proved. The remaining steps are named propositions in the statement file: Poisson summation for the whole class, the scaled Poisson formula, and GW-EF, the link from Poisson summation to the Weil form. - The Riemann Hypothesis, one of the framework's targets. Holographically, the
functional equation says the hologram is symmetric, and RH says it is also positive
(reflection-positive across the equator). The finite core is proved:
GppHolographicRP.reflectionForm_nonneg_iffshows that positivity across the equator holds iff there are no off-equator mirror pairs, andGppWeilCriterion.rh_of_weil_pairedForm_nonnegreduces RH to Weil positivity. The finite windows of that positivity are certified: the Schur sector form is positive for sectors K = 2 to 18, and the first two zero orbitals close the K = 5 diagonal in exact rational arithmetic (the AUG2 certificate, an independent tail sum over 138,067 further zeros confirms the closure is a fact about the zeros). The current target is the positivity stated as unitarity of the global S-matrix; its open statement is named (SPEC), with GW-EF as the named link to the Weil form. - Physics on the same involution. Half-flip / CPT and time reversal, Majorana
conditions, complete positivity, Grassmannian
Gr(2,4)geometry, and a Standard Model sector:StandardModel/,QuantumGravity/, and the smaller sectors in the layout table below. Physics inputs enter as explicit, named hypotheses. - No-go theorems. Routes that fail are kept as theorems pinning down exactly where they
fail, for example the Fisher-zero log-concavity no-go (
FisherZeroLogConcavityNoGo.lean). They live inNoGo/and in the layers they close off. - Upstream staging.
Upstream/holds files written to Mathlib conventions for contribution to Mathlib: the Weierstrass and entire-function order-of-growth cluster, the two adele files of the Tate layer, and the two-vector projection formulas. The first contribution is open as mathlib4 PR #45825; the rest are staged.
- Every theorem is kernel-checked, with zero
sorryand zero custom axioms;scripts/check_axioms.leanaudits the count. - Steps still being built are
open_…stubs:theorem open_foo : True := trivial, with a doc comment naming the exact mathematics that will fill it. Theopen_prefix is CI-checked, and a stub is retired by proving it.
Each point below is a command or a file.
- The epistemics are mechanical.
scripts/holds the gates CI runs:check_sorries.py,check_stub_naming.pyandcheck_vacuity.py(every declaration carries real content),check_blueprint_refs.py(every\lean{}in the blueprint must name a real declaration),check_landing_claims.py(the public page matches the tree),check_import_graph.py, andcheck_axioms.lean(#print axiomson the flagship theorems). Each script's header records the audit that produced it. - The Riemann Hypothesis is stated as an exact equivalence.
rh_iff_weil_pairedForm_nonneg(RH iff a positivity statement) andGppHolographicRP.reflectionForm_nonneg_iff(positivity across the equator iff every zero pairs on the equator) are proved, and the positivity step is named. - No-go results are first-class output, kept as theorems, e.g.
FisherZeroLogConcavityNoGo.lean. - The
open_…stubs are counted and named, so the stub ledger is the live map of what is being built next. - Dates are on everything.
docs/holds dated working notes andarchive/holds earlier surveys; the current status is this file and the blueprint.
The current module, stub, and axiom counts are published on the
blueprint and kept in sync by scripts/sync_published_counts.py.
To reproduce them:
grep -rn "^\s*sorry\s*$" --include="*.lean" GppVerify/ | wc -l
grep -rn "^axiom " --include="*.lean" GppVerify/ | wc -l
grep -rn ": True := trivial" --include="*.lean" GppVerify/ | wc -lcurl -sSfL https://github.com/leanprover/elan/releases/latest/download/elan-x86_64-unknown-linux-gnu.tar.gz | tar xz
./elan-init -y
lake exe cache get
lake buildBlueprint: pip install leanblueprint && cd blueprint && leanblueprint build. The output is
blueprint/web/index.html.
| Directory | Files | Contents |
|---|---|---|
GppVerify/RiemannHypothesis/ |
585 | the RH programme: functional equation, Tate layer, explicit formula, Weil positivity reductions, certificates |
GppVerify/CelestialHolography/ |
329 | the celestial side: shadow transform, principal series, Penrose and ambitwistor geometry |
GppVerify/StandardModel/ |
132 | the Standard Model sector: CPT and orientation, charge, Pati–Salam |
GppVerify/QuantumGravity/ |
53 | quantum gravity: causal diamonds, equator thermodynamics, Fisher geometry |
GppVerify/NoGo/ |
23 | no-go theorems: routes closed, with the failing step located |
GppVerify/GeneralRelativity/ |
20 | causal diamonds and boundary charges in GR form |
GppVerify/NumberTheory/ |
19 | zeta values, Euler sums, BSD point counts, shadow Euler identities |
GppVerify/ThreadWeilParity/ |
16 | the Weil parity thread |
GppVerify/Upstream/ |
13 | Mathlib staging |
GppVerify/Analysis/ |
5 | Landau and Pringsheim theory for Dirichlet series |
GppVerify/Cosmology/ |
4 | cosmological sectors |
GppVerify/QuantumInformation/ |
4 | quantum information sectors |
GppVerify/QuantumFieldTheory/ |
3 | QFT sectors |
GppVerify/ThreadS/, ThreadHT/, ThreadWeilSemibound/ |
5 | self-contained research threads |
GppVerify/Certificates/, StringTheory/, YangMills/ |
3 | single-file sectors |
discovery/ |
exploratory notes that sit next to the Lean threads they feed | |
docs/ |
dated working notes, and the Tate guide | |
blueprint/ |
the technical blueprint source | |
scripts/ |
the CI gates | |
archive/ |
earlier surveys, including the dependency map |