Pith. sign in

IndisputableMonolith.Verification.Item8ClosureTarget

IndisputableMonolith/Verification/Item8ClosureTarget.lean · 1069 lines · 102 declarations

show as:
view math explainer →

open module explainer GitHub source

Explainer status: pending

   1import Mathlib
   2import IndisputableMonolith.Constants
   3import IndisputableMonolith.Constants.Alpha
   4import IndisputableMonolith.Physics.RunningCouplings
   5
   6/-!
   7# Item 8 Closure Target — Unified Sub-Leading Mass Formula
   8
   9This module develops the smallest precise theorem framework that would close the
  10open quark sub-leading correction item (Item 8) and make the all-sector
  11generalization falsifiable.
  12
  13## Summary of Results
  14
  15### Proved (sorry-free)
  16
  171. **`consistency_of_ratioFamily`**: The sign-split family forces
  18   `gen12·s₁₂ + gen23·s₂₃ = 0` — a structural rigidity that PDG data violates.
  19
  202. **`etaFromData` + `etaL_gen12_identity` + `etaL_gen23_identity`**:
  21   Closed-form η formula absorbs the consistency violation:
  22   `η = (g₁₂·s₁₂ + g₂₃·s₂₃) / (ln(s₁₂/s₂₃) · (g₁₂·s₁₂ − g₂₃·s₂₃))`
  23   with algebraic identities `1 + η·L = 2·g₁₂·s₁₂/D` and `1 − η·L = −2·g₂₃·s₂₃/D`.
  24
  253. **`refinedFamily_neg_solvable` / `refinedFamily_pos_solvable`**:
  26   For any residual pair with nonzero gen12, non-degenerate cross-difference,
  27   distinct steps, and nonzero coupling, the refined family has a solution.
  28   Constructive proof via explicit coefficient formulas.
  29
  304. **`refinedFamily_neg_unique` / `refinedFamily_pos_unique`**:
  31   If two coefficient sets produce identical refined-family outputs on a given
  32   signature, their active coefficient `c` and `eta` must agree.
  33   Proof via `calc`+`ring` to cancel common factors, then `linarith` for the
  34   addition step and `mul_left_cancel₀`/`mul_right_cancel₀` for cancellation.
  35   Combined with solvability (item 3), this gives **`∃!` for each sector**.
  36
  375. **`refined_neg_sector_closure`**: Full `∃!` for any `.neg` sector:
  38   there exist unique `(c, η)` matching the data, and any solution (with
  39   arbitrary `cPos`) must have the same `(c, η)`.  Combines items 3–4 with
  40   a helper `neg_cPos_irrelevant` showing `cPos` is unused for `.neg`.
  41
  426. **`signClass_collapse_to_refined`**: Sign-class family collapses to the
  43   refined family when both η values agree.
  44
  45### Documented obstructions
  46
  47* **Universal η**: The 3-parameter refined family (universal η) is
  48  unsatisfiable with mixed-scheme PDG quark data: η_up = −2.72 vs η_dn = −0.88
  49  (3.1× disagreement).
  50
  51* **Mixed-scheme artifact**: LO QCD running reveals that the "universal sign
  52  pattern" (δ₁₂ > 0, δ₂₃ < 0) is partly an artifact of comparing masses at
  53  different renormalization scales.  The c/u SDGT step of 13 is nearly exact
  54  at a common scale (δ₁₂ ≈ −0.01).
  55
  56* **Lepton benchmark**: Leptons — with scheme-free pole masses — give
  57  perturbative η = +0.065 (|η·L| ≈ 4%), validating the family form.
  58
  59### Strategy
  60
  61The lepton-anchored approach fixes η and cNeg from the two scheme-free lepton
  62equations, then predicts anchor-scale quark residuals as genuine out-of-sample
  63tests.  The `RunningCouplings` module now provides LO mass transport with
  64threshold matching (`transport_mass_through`); quark closure awaits NLO
  65refinement of the running to reduce residual scheme dependence.
  66
  67## Interface
  68
  69* A sign class induced by `B_pow`
  70* The ordered SDGT step pair `(step12, step23)` — sector-specific rung spacings
  71  from the Q₃ cube decomposition
  72* One coupling scalar `kappa`
  73
  74## Concrete data (§ ConcreteInstantiation)
  75
  76* PDG 2022 quark masses at mixed reference scales
  77  (u,d,s at 2 GeV MS-bar; c at m_c; b at m_b; t pole)
  78* Charged-lepton pole masses (scheme-free)
  79* α_s = 2/17 from wallpaper-group fraction (proved in StrongForce)
  80* Rung-unit residuals: δ = log_φ(observed ratio) − SDGT step
  81
  82### Approximate residual values (mixed-scheme PDG)
  83
  84| Sector | gen12 step | δ₁₂ (rungs) | gen23 step | δ₂₃ (rungs) | η     |
  85|--------|-----------|-------------|-----------|-------------|-------|
  86| Up     | 13        | +0.25       | 11        | −0.79       | −2.72 |
  87| Down   | 6         | +0.23       | 8         | −0.10       | −0.88 |
  88| Lepton | 11        | +0.08       | 6         | −0.13       | +0.06 |
  89
  90Lepton η is perturbative (4%); quark η values are inflated by mixed-scale
  91PDG artifacts, especially the t/c ratio (pole mass vs MS-bar at m_c).
  92-/
  93
  94namespace IndisputableMonolith
  95namespace Verification
  96namespace Item8ClosureTarget
  97
  98/-- Sign class induced by the sign of `B_pow`. -/
  99inductive BpowSign | neg | pos
 100  deriving DecidableEq, Repr
 101
 102/-- Minimal structural signature for a sector's sub-leading correction law.
 103    `step12` and `step23` are the sector-specific SDGT rung spacings
 104    (cube-cell counts from the Q₃ decomposition). -/
 105structure ResidualSignature where
 106  sign : BpowSign
 107  step12 : ℕ
 108  step23 : ℕ
 109  coupling : ℝ
 110  step12_pos : 0 < step12
 111  step23_pos : 0 < step23
 112
 113/-- Two sub-leading rung corrections: generation `1 → 2` and `2 → 3`. -/
 114@[ext]
 115structure ResidualPair where
 116  gen12 : ℝ
 117  gen23 : ℝ
 118
 119/-- The two global coefficients allowed by the smallest sign-split candidate family. -/
 120structure RatioFamilyCoeffs where
 121  cNeg : ℝ
 122  cPos : ℝ
 123
 124/-- Negative-`B_pow` lepton signature, using the derived SDGT pair `(11, 6)`. -/
 125def leptonSignature (kappa : ℝ) : ResidualSignature where
 126  sign := .neg
 127  step12 := 11
 128  step23 := 6
 129  coupling := kappa
 130  step12_pos := by decide
 131  step23_pos := by decide
 132
 133/-- Negative-`B_pow` up-quark signature, using the derived SDGT pair `(13, 11)`. -/
 134def upQuarkSignature (kappa : ℝ) : ResidualSignature where
 135  sign := .neg
 136  step12 := 13
 137  step23 := 11
 138  coupling := kappa
 139  step12_pos := by decide
 140  step23_pos := by decide
 141
 142/-- Positive-`B_pow` down-quark signature, using the derived SDGT pair `(6, 8)`. -/
 143def downQuarkSignature (kappa : ℝ) : ResidualSignature where
 144  sign := .pos
 145  step12 := 6
 146  step23 := 8
 147  coupling := kappa
 148  step12_pos := by decide
 149  step23_pos := by decide
 150
 151/-- Sign-split ratio family: the smallest closed-form candidate using only
 152
 1531. `sign(B_pow)` → selects `cNeg` or `cPos`
 1542. the ordered SDGT step pair → weights the correction via step fraction
 1553. one coupling scalar κ
 156
 157The gen-1→2 correction inherits the sign of `c`; the gen-2→3 correction
 158inherits `−c`.  This captures the universal empirical pattern δ₁₂ > 0,
 159δ₂₃ < 0 when `c > 0`. -/
 160noncomputable def ratioFamily (coeffs : RatioFamilyCoeffs)
 161    (sig : ResidualSignature) : ResidualPair :=
 162  let c :=
 163    match sig.sign with
 164    | .neg => coeffs.cNeg
 165    | .pos => coeffs.cPos
 166  let total := (sig.step12 : ℝ) + (sig.step23 : ℝ)
 167  { gen12 := c * sig.coupling * (sig.step23 : ℝ) / total
 168  , gen23 := -(c * sig.coupling * (sig.step12 : ℝ) / total)
 169  }
 170
 171/-- The exact prediction produced by the candidate family for any signature. -/
 172noncomputable def predictedResiduals (coeffs : RatioFamilyCoeffs)
 173    (sig : ResidualSignature) : ResidualPair :=
 174  ratioFamily coeffs sig
 175
 176/-- Equal structural signatures give equal predictions.  This is the precise
 177no-extra-knob property needed for later out-of-sample tests. -/
 178theorem same_signature_same_prediction
 179    (coeffs : RatioFamilyCoeffs) {sig1 sig2 : ResidualSignature}
 180    (h : sig1 = sig2) :
 181    predictedResiduals coeffs sig1 = predictedResiduals coeffs sig2 := by
 182  subst h; rfl
 183
 184/-- Smallest precise target that would close Item 8.
 185
 186`upExact` and `downExact` are the exact quark residual pairs once derived.
 187Proving this proposition would do two things:
 188
 1891. close Item 8 by fixing both quark residual pairs inside one closed family
 1902. freeze the two global coefficients `(cNeg, cPos)`, making every later
 191   lepton / genetic / theta instantiation an out-of-sample test rather than a
 192   new fit
 193-/
 194def item8ClosureTarget
 195    (upExact downExact : ResidualPair)
 196    (kappaUp kappaDown : ℝ) : Prop :=
 197  ∃! coeffs : RatioFamilyCoeffs,
 198    predictedResiduals coeffs (upQuarkSignature kappaUp) = upExact ∧
 199    predictedResiduals coeffs (downQuarkSignature kappaDown) = downExact
 200
 201/-! ## Structural Obstruction of the Sign-Split Family
 202
 203The sign-split family satisfies a rigid algebraic constraint: for any
 204signature, the cross-product `gen12 · step12 + gen23 · step23 = 0`.
 205This is a **necessary condition** for the family to match any given
 206residual pair.  Numerically:
 207
 208| Sector | gen12·s₁₂ + gen23·s₂₃ | violation |
 209|--------|----------------------|-----------|
 210| Lepton |  +0.066              |  3.9%     |
 211| Down   |  +0.546              | 25.3%     |
 212| Up     |  −5.439              | 45.4%     |
 213
 214The lepton data nearly satisfies the condition (scheme-free pole masses);
 215the quark violations are larger, partly because PDG values mix scales.
 216The up-quark t/c ratio (pole vs MS-bar at m_c) is the main outlier.
 217-/
 218
 219theorem consistency_of_ratioFamily
 220    (coeffs : RatioFamilyCoeffs) (sig : ResidualSignature) :
 221    (ratioFamily coeffs sig).gen12 * (sig.step12 : ℝ) +
 222    (ratioFamily coeffs sig).gen23 * (sig.step23 : ℝ) = 0 := by
 223  obtain ⟨sign, s12, s23, κ, hs12, hs23⟩ := sig
 224  unfold ratioFamily
 225  dsimp only [ResidualPair.gen12, ResidualPair.gen23]
 226  have htotal : (s12 : ℝ) + (s23 : ℝ) ≠ 0 := by
 227    have : (0 : ℝ) < (s12 : ℝ) := Nat.cast_pos.mpr hs12
 228    have : (0 : ℝ) < (s23 : ℝ) := Nat.cast_pos.mpr hs23
 229    linarith
 230  cases sign <;> simp only [] <;> field_simp [htotal] <;> ring
 231
 232/-- If the sign-split family matches a residual pair, that pair must satisfy
 233    the cross-product identity.  Contrapositive: if the identity fails,
 234    no coefficients can make the family fit. -/
 235theorem consistency_necessary
 236    (coeffs : RatioFamilyCoeffs) (sig : ResidualSignature) (exact : ResidualPair)
 237    (h : predictedResiduals coeffs sig = exact) :
 238    exact.gen12 * (sig.step12 : ℝ) + exact.gen23 * (sig.step23 : ℝ) = 0 := by
 239  rw [← h]; simp only [predictedResiduals]; exact consistency_of_ratioFamily coeffs sig
 240
 241/-! ## Refined Family: Log-Asymmetry Correction
 242
 243Since the sign-split family's structural constraint is violated by the data,
 244we introduce a **single universal parameter** `η` that breaks the rigid
 245ratio `gen12/gen23 = −step23/step12`.  The correction is multiplicative and
 246logarithmic in the step ratio:
 247
 248    gen12 = c · κ · (s₂₃/total) · (1 + η · ln(s₁₂/s₂₃))
 249    gen23 = −c · κ · (s₁₂/total) · (1 − η · ln(s₁₂/s₂₃))
 250
 251With 3 parameters (cNeg, cPos, η) and 4 quark equations, the system is
 252overdetermined by 1, making the `∃!` condition a genuine test (not a tautology).
 253-/
 254
 255/-- Three global coefficients: two sign-class amplitudes plus a universal
 256    log-asymmetry that modulates the gen12/gen23 ratio. -/
 257structure RefinedCoeffs where
 258  cNeg : ℝ
 259  cPos : ℝ
 260  eta  : ℝ
 261
 262noncomputable def refinedFamily (coeffs : RefinedCoeffs)
 263    (sig : ResidualSignature) : ResidualPair :=
 264  let c :=
 265    match sig.sign with
 266    | .neg => coeffs.cNeg
 267    | .pos => coeffs.cPos
 268  let s12 := (sig.step12 : ℝ)
 269  let s23 := (sig.step23 : ℝ)
 270  let total := s12 + s23
 271  let logAsym := Real.log (s12 / s23)
 272  { gen12 := c * sig.coupling * s23 / total * (1 + coeffs.eta * logAsym)
 273  , gen23 := -(c * sig.coupling * s12 / total * (1 - coeffs.eta * logAsym))
 274  }
 275
 276noncomputable def refinedPrediction (coeffs : RefinedCoeffs)
 277    (sig : ResidualSignature) : ResidualPair :=
 278  refinedFamily coeffs sig
 279
 280/-- At η = 0, the refined family collapses to the original sign-split family. -/
 281theorem refined_at_eta_zero (cN cP : ℝ) (sig : ResidualSignature) :
 282    refinedFamily ⟨cN, cP, 0⟩ sig =
 283    ratioFamily ⟨cN, cP⟩ sig := by
 284  simp only [refinedFamily, ratioFamily, zero_mul, add_zero, sub_zero, mul_one]
 285
 286def refinedItem8ClosureTarget
 287    (upExact downExact : ResidualPair)
 288    (kappaUp kappaDown : ℝ) : Prop :=
 289  ∃! coeffs : RefinedCoeffs,
 290    refinedPrediction coeffs (upQuarkSignature kappaUp) = upExact ∧
 291    refinedPrediction coeffs (downQuarkSignature kappaDown) = downExact
 292
 293/-! ## Closed-Form η Solution
 294
 295For any sector with residual pair `(g₁₂, g₂₃)` and SDGT steps `(s₁₂, s₂₃)`,
 296the log-asymmetry parameter η that makes the refined family match the data
 297exactly has a unique closed-form expression:
 298
 299    η = (g₁₂·s₁₂ + g₂₃·s₂₃) / (ln(s₁₂/s₂₃) · (g₁₂·s₁₂ − g₂₃·s₂₃))
 300
 301The numerator is exactly the consistency violation of the sign-split family
 302(which forces `g₁₂·s₁₂ + g₂₃·s₂₃ = 0`), and the denominator is the
 303log-asymmetry times the cross-difference.
 304
 305### Key algebraic identities
 306
 307When η has this value, the multiplicative correction factors simplify:
 308
 309    1 + η·L = 2·g₁₂·s₁₂ / (g₁₂·s₁₂ − g₂₃·s₂₃)
 310    1 − η·L = −2·g₂₃·s₂₃ / (g₁₂·s₁₂ − g₂₃·s₂₃)
 311
 312where `L = ln(s₁₂/s₂₃)`.  These identities reduce the solution for `c`
 313to a single division, and the entire system closes algebraically.
 314
 315### Obstruction for universal η
 316
 317The formula yields different η values per sector with current PDG data:
 318
 319| Sector | η       | |η·L|  | Perturbative? |
 320|--------|---------|--------|---------------|
 321| Lepton | +0.065  | 0.039  | ✓ (4%)        |
 322| Down   | −0.879  | 0.253  | marginal      |
 323| Up     | −2.720  | 0.454  | ✗ (flips sign)|
 324
 325The 3-parameter refined family (universal η) is therefore UNSATISFIABLE
 326with current mixed-scheme PDG quark data.  Leptons — with scheme-free
 327pole masses — give a perturbative η, validating the family structure.
 328The quark η inflation is attributed to mixed-scale PDG artifacts
 329(especially the t/c ratio: pole vs MS-bar at m_c).
 330-/
 331
 332/-- Closed-form η from residual data: absorbs the consistency violation
 333    of the sign-split family into a log-asymmetry correction.
 334    Well-defined when `s₁₂ ≠ s₂₃` and `g₁₂·s₁₂ ≠ g₂₃·s₂₃`. -/
 335noncomputable def etaFromData (g12 g23 s12 s23 : ℝ) : ℝ :=
 336  (g12 * s12 + g23 * s23) / (Real.log (s12 / s23) * (g12 * s12 - g23 * s23))
 337
 338/-- When η is computed from the data, `1 + η·L` simplifies to
 339    `2·g₁₂·s₁₂ / (g₁₂·s₁₂ − g₂₃·s₂₃)`.  This is the identity that
 340    makes the gen-1→2 equation close algebraically. -/
 341theorem etaL_gen12_identity (g12 g23 s12 s23 : ℝ)
 342    (hD : g12 * s12 - g23 * s23 ≠ 0)
 343    (hL : Real.log (s12 / s23) ≠ 0) :
 344    1 + etaFromData g12 g23 s12 s23 * Real.log (s12 / s23) =
 345    2 * g12 * s12 / (g12 * s12 - g23 * s23) := by
 346  unfold etaFromData
 347  field_simp
 348  ring
 349
 350/-- Companion identity: `1 − η·L = −2·g₂₃·s₂₃ / (g₁₂·s₁₂ − g₂₃·s₂₃)`.
 351    This closes the gen-2→3 equation. -/
 352theorem etaL_gen23_identity (g12 g23 s12 s23 : ℝ)
 353    (hD : g12 * s12 - g23 * s23 ≠ 0)
 354    (hL : Real.log (s12 / s23) ≠ 0) :
 355    1 - etaFromData g12 g23 s12 s23 * Real.log (s12 / s23) =
 356    -(2 * g23 * s23) / (g12 * s12 - g23 * s23) := by
 357  unfold etaFromData
 358  field_simp
 359  ring
 360
 361/-- The consistency violation of the sign-split family is EXACTLY
 362    what η absorbs: `g₁₂·s₁₂ + g₂₃·s₂₃ = η · L · (g₁₂·s₁₂ − g₂₃·s₂₃)`.
 363    This is the fundamental equation relating the log-asymmetry parameter
 364    to the data's departure from the rigid `gen12·s12 + gen23·s23 = 0` law. -/
 365theorem eta_absorbs_consistency (g12 g23 s12 s23 : ℝ)
 366    (hD : g12 * s12 - g23 * s23 ≠ 0)
 367    (hL : Real.log (s12 / s23) ≠ 0) :
 368    g12 * s12 + g23 * s23 =
 369    etaFromData g12 g23 s12 s23 * Real.log (s12 / s23) *
 370      (g12 * s12 - g23 * s23) := by
 371  unfold etaFromData
 372  field_simp
 373
 374/-- **SINGLE-SECTOR EXISTENCE**: For any residual pair with nonzero gen12,
 375    non-degenerate cross-difference, distinct steps, and nonzero coupling,
 376    the refined family has a solution.  The proof constructs explicit
 377    coefficients using `etaFromData` and verifies both components
 378    via the `etaL_gen12/gen23_identity` theorems. -/
 379theorem refinedFamily_neg_solvable
 380    (g12 g23 : ℝ) (s12 s23 : ℕ) (κ : ℝ)
 381    (hs12 : 0 < s12) (hs23 : 0 < s23)
 382    (hκ : κ ≠ 0)
 383    (hg12 : g12 ≠ 0)
 384    (hD : g12 * (s12 : ℝ) - g23 * (s23 : ℝ) ≠ 0)
 385    (hL : Real.log ((s12 : ℝ) / (s23 : ℝ)) ≠ 0) :
 386    ∃ coeffs : RefinedCoeffs,
 387      refinedFamily coeffs ⟨.neg, s12, s23, κ, hs12, hs23⟩ =
 388        ⟨g12, g23⟩ := by
 389  have hs23_ne : (s23 : ℝ) ≠ 0 := Nat.cast_ne_zero.mpr (by omega)
 390  have hs12_ne : (s12 : ℝ) ≠ 0 := Nat.cast_ne_zero.mpr (by omega)
 391  have h12 := etaL_gen12_identity g12 g23 (s12 : ℝ) (s23 : ℝ) hD hL
 392  have h23 := etaL_gen23_identity g12 g23 (s12 : ℝ) (s23 : ℝ) hD hL
 393  have h_corr_ne : 1 + etaFromData g12 g23 (s12 : ℝ) (s23 : ℝ) *
 394      Real.log ((s12 : ℝ) / (s23 : ℝ)) ≠ 0 := by
 395    rw [h12]
 396    exact div_ne_zero (mul_ne_zero (mul_ne_zero (by norm_num) hg12) hs12_ne) hD
 397  refine ⟨⟨g12 * ((s12 : ℝ) + (s23 : ℝ)) /
 398    (κ * (s23 : ℝ) * (1 + etaFromData g12 g23 (s12 : ℝ) (s23 : ℝ) *
 399      Real.log ((s12 : ℝ) / (s23 : ℝ)))),
 400    0, etaFromData g12 g23 (s12 : ℝ) (s23 : ℝ)⟩, ?_⟩
 401  have hD_swap : g12 * (s12 : ℝ) - (s23 : ℝ) * g23 ≠ 0 := by
 402    rw [show g12 * (s12 : ℝ) - (s23 : ℝ) * g23 =
 403        g12 * (s12 : ℝ) - g23 * (s23 : ℝ) from by ring]; exact hD
 404  have hD_comm : -(g23 * (s23 : ℝ)) + g12 * (s12 : ℝ) ≠ 0 := by
 405    intro h; exact hD (by linarith)
 406  simp only [refinedFamily]
 407  ext
 408  · -- gen12: direct cancellation of cNeg · κ · s23 / total · (1 + η·L)
 409    dsimp [ResidualPair.gen12]
 410    field_simp
 411  · -- gen23: substitute correction factors, then cancel D/D
 412    dsimp [ResidualPair.gen23]
 413    rw [h12, h23]
 414    field_simp [hD_comm, hD_swap]
 415
 416/-- The positive-sign variant: same existence for `BpowSign.pos` sectors. -/
 417theorem refinedFamily_pos_solvable
 418    (g12 g23 : ℝ) (s12 s23 : ℕ) (κ : ℝ)
 419    (hs12 : 0 < s12) (hs23 : 0 < s23)
 420    (hκ : κ ≠ 0)
 421    (hg12 : g12 ≠ 0)
 422    (hD : g12 * (s12 : ℝ) - g23 * (s23 : ℝ) ≠ 0)
 423    (hL : Real.log ((s12 : ℝ) / (s23 : ℝ)) ≠ 0) :
 424    ∃ coeffs : RefinedCoeffs,
 425      refinedFamily coeffs ⟨.pos, s12, s23, κ, hs12, hs23⟩ =
 426        ⟨g12, g23⟩ := by
 427  have hs23_ne : (s23 : ℝ) ≠ 0 := Nat.cast_ne_zero.mpr (by omega)
 428  have hs12_ne : (s12 : ℝ) ≠ 0 := Nat.cast_ne_zero.mpr (by omega)
 429  have h12 := etaL_gen12_identity g12 g23 (s12 : ℝ) (s23 : ℝ) hD hL
 430  have h23 := etaL_gen23_identity g12 g23 (s12 : ℝ) (s23 : ℝ) hD hL
 431  have h_corr_ne : 1 + etaFromData g12 g23 (s12 : ℝ) (s23 : ℝ) *
 432      Real.log ((s12 : ℝ) / (s23 : ℝ)) ≠ 0 := by
 433    rw [h12]
 434    exact div_ne_zero (mul_ne_zero (mul_ne_zero (by norm_num) hg12) hs12_ne) hD
 435  have hD_swap : g12 * (s12 : ℝ) - (s23 : ℝ) * g23 ≠ 0 := by
 436    rw [show g12 * (s12 : ℝ) - (s23 : ℝ) * g23 =
 437        g12 * (s12 : ℝ) - g23 * (s23 : ℝ) from by ring]; exact hD
 438  have hD_comm : -(g23 * (s23 : ℝ)) + g12 * (s12 : ℝ) ≠ 0 := by
 439    intro h; exact hD (by linarith)
 440  refine ⟨⟨0, g12 * ((s12 : ℝ) + (s23 : ℝ)) /
 441    (κ * (s23 : ℝ) * (1 + etaFromData g12 g23 (s12 : ℝ) (s23 : ℝ) *
 442      Real.log ((s12 : ℝ) / (s23 : ℝ)))),
 443    etaFromData g12 g23 (s12 : ℝ) (s23 : ℝ)⟩, ?_⟩
 444  simp only [refinedFamily]
 445  ext
 446  · dsimp [ResidualPair.gen12]
 447    field_simp
 448  · dsimp [ResidualPair.gen23]
 449    rw [h12, h23]
 450    field_simp [hD_comm, hD_swap]
 451
 452/-- **SINGLE-SECTOR UNIQUENESS (neg)**: If two coefficient sets (with sign
 453    class `.neg`) produce identical refined-family outputs on a given
 454    signature, their active coefficient `cNeg` and `eta` must agree.
 455    Combined with `refinedFamily_neg_solvable`, this gives `∃!`.
 456
 457    Proof strategy: extract gen12/gen23 component equations, cancel the
 458    common factor `κ·s/total` from each, then add the two simplified
 459    equations to get `c₁ = c₂`, and substitute back to get `eta₁ = eta₂`. -/
 460theorem refinedFamily_neg_unique
 461    (c₁ c₂ p₁ p₂ eta₁ eta₂ : ℝ)
 462    (s12 s23 : ℕ) (κ : ℝ)
 463    (hs12 : 0 < s12) (hs23 : 0 < s23)
 464    (hκ : κ ≠ 0) (hc1 : c₁ ≠ 0)
 465    (hL : Real.log ((s12 : ℝ) / (s23 : ℝ)) ≠ 0)
 466    (h : refinedFamily ⟨c₁, p₁, eta₁⟩ ⟨.neg, s12, s23, κ, hs12, hs23⟩ =
 467         refinedFamily ⟨c₂, p₂, eta₂⟩ ⟨.neg, s12, s23, κ, hs12, hs23⟩) :
 468    c₁ = c₂ ∧ eta₁ = eta₂ := by
 469  have hs23_ne : (s23 : ℝ) ≠ 0 := Nat.cast_ne_zero.mpr (by omega)
 470  have hs12_ne : (s12 : ℝ) ≠ 0 := Nat.cast_ne_zero.mpr (by omega)
 471  have htotal_ne : (s12 : ℝ) + (s23 : ℝ) ≠ 0 := ne_of_gt (by positivity)
 472  -- Extract gen12 and gen23 component equations from the ResidualPair equality
 473  have hg12 := congr_arg ResidualPair.gen12 h
 474  have hg23 := congr_arg ResidualPair.gen23 h
 475  dsimp only [refinedFamily] at hg12 hg23
 476  -- hg23 has negation on both sides; strip it
 477  replace hg23 := neg_inj.mp hg23
 478  -- Cancel κ·s23/total from gen12 equation via calc + ring
 479  have h1 : c₁ * (1 + eta₁ * Real.log (↑s12 / ↑s23)) =
 480             c₂ * (1 + eta₂ * Real.log (↑s12 / ↑s23)) := by
 481    have hf : κ * (↑s23 : ℝ) / ((↑s12 : ℝ) + ↑s23) ≠ 0 :=
 482      div_ne_zero (mul_ne_zero hκ hs23_ne) htotal_ne
 483    exact mul_right_cancel₀ hf (show
 484      c₁ * (1 + eta₁ * Real.log (↑s12 / ↑s23)) * (κ * ↑s23 / (↑s12 + ↑s23)) =
 485      c₂ * (1 + eta₂ * Real.log (↑s12 / ↑s23)) * (κ * ↑s23 / (↑s12 + ↑s23)) from by
 486        calc _ = c₁ * κ * ↑s23 / (↑s12 + ↑s23) *
 487                   (1 + eta₁ * Real.log (↑s12 / ↑s23)) := by ring
 488             _ = c₂ * κ * ↑s23 / (↑s12 + ↑s23) *
 489                   (1 + eta₂ * Real.log (↑s12 / ↑s23)) := hg12
 490             _ = _ := by ring)
 491  -- Cancel κ·s12/total from gen23 equation
 492  have h2 : c₁ * (1 - eta₁ * Real.log (↑s12 / ↑s23)) =
 493             c₂ * (1 - eta₂ * Real.log (↑s12 / ↑s23)) := by
 494    have hf : κ * (↑s12 : ℝ) / ((↑s12 : ℝ) + ↑s23) ≠ 0 :=
 495      div_ne_zero (mul_ne_zero hκ hs12_ne) htotal_ne
 496    exact mul_right_cancel₀ hf (show
 497      c₁ * (1 - eta₁ * Real.log (↑s12 / ↑s23)) * (κ * ↑s12 / (↑s12 + ↑s23)) =
 498      c₂ * (1 - eta₂ * Real.log (↑s12 / ↑s23)) * (κ * ↑s12 / (↑s12 + ↑s23)) from by
 499        calc _ = c₁ * κ * ↑s12 / (↑s12 + ↑s23) *
 500                   (1 - eta₁ * Real.log (↑s12 / ↑s23)) := by ring
 501             _ = c₂ * κ * ↑s12 / (↑s12 + ↑s23) *
 502                   (1 - eta₂ * Real.log (↑s12 / ↑s23)) := hg23
 503             _ = _ := by ring)
 504  -- Add h1 and h2: eta terms cancel, leaving 2·c₁ = 2·c₂
 505  have hc : c₁ = c₂ := by linarith
 506  -- Substitute c₁ = c₂ into h1, cancel c₁ (≠ 0) then L (≠ 0)
 507  have heta : eta₁ = eta₂ := by
 508    rw [← hc] at h1
 509    have := mul_left_cancel₀ hc1 h1
 510    have : eta₁ * Real.log (↑s12 / ↑s23) = eta₂ * Real.log (↑s12 / ↑s23) := by linarith
 511    exact mul_right_cancel₀ hL this
 512  exact ⟨hc, heta⟩
 513
 514/-- **SINGLE-SECTOR UNIQUENESS (pos)**: Symmetric variant for `.pos` sign class. -/
 515theorem refinedFamily_pos_unique
 516    (n₁ n₂ c₁ c₂ eta₁ eta₂ : ℝ)
 517    (s12 s23 : ℕ) (κ : ℝ)
 518    (hs12 : 0 < s12) (hs23 : 0 < s23)
 519    (hκ : κ ≠ 0) (hc1 : c₁ ≠ 0)
 520    (hL : Real.log ((s12 : ℝ) / (s23 : ℝ)) ≠ 0)
 521    (h : refinedFamily ⟨n₁, c₁, eta₁⟩ ⟨.pos, s12, s23, κ, hs12, hs23⟩ =
 522         refinedFamily ⟨n₂, c₂, eta₂⟩ ⟨.pos, s12, s23, κ, hs12, hs23⟩) :
 523    c₁ = c₂ ∧ eta₁ = eta₂ := by
 524  have hs23_ne : (s23 : ℝ) ≠ 0 := Nat.cast_ne_zero.mpr (by omega)
 525  have hs12_ne : (s12 : ℝ) ≠ 0 := Nat.cast_ne_zero.mpr (by omega)
 526  have htotal_ne : (s12 : ℝ) + (s23 : ℝ) ≠ 0 := ne_of_gt (by positivity)
 527  have hg12 := congr_arg ResidualPair.gen12 h
 528  have hg23 := congr_arg ResidualPair.gen23 h
 529  dsimp only [refinedFamily] at hg12 hg23
 530  replace hg23 := neg_inj.mp hg23
 531  have h1 : c₁ * (1 + eta₁ * Real.log (↑s12 / ↑s23)) =
 532             c₂ * (1 + eta₂ * Real.log (↑s12 / ↑s23)) := by
 533    have hf : κ * (↑s23 : ℝ) / ((↑s12 : ℝ) + ↑s23) ≠ 0 :=
 534      div_ne_zero (mul_ne_zero hκ hs23_ne) htotal_ne
 535    exact mul_right_cancel₀ hf (show
 536      c₁ * (1 + eta₁ * Real.log (↑s12 / ↑s23)) * (κ * ↑s23 / (↑s12 + ↑s23)) =
 537      c₂ * (1 + eta₂ * Real.log (↑s12 / ↑s23)) * (κ * ↑s23 / (↑s12 + ↑s23)) from by
 538        calc _ = c₁ * κ * ↑s23 / (↑s12 + ↑s23) *
 539                   (1 + eta₁ * Real.log (↑s12 / ↑s23)) := by ring
 540             _ = c₂ * κ * ↑s23 / (↑s12 + ↑s23) *
 541                   (1 + eta₂ * Real.log (↑s12 / ↑s23)) := hg12
 542             _ = _ := by ring)
 543  have h2 : c₁ * (1 - eta₁ * Real.log (↑s12 / ↑s23)) =
 544             c₂ * (1 - eta₂ * Real.log (↑s12 / ↑s23)) := by
 545    have hf : κ * (↑s12 : ℝ) / ((↑s12 : ℝ) + ↑s23) ≠ 0 :=
 546      div_ne_zero (mul_ne_zero hκ hs12_ne) htotal_ne
 547    exact mul_right_cancel₀ hf (show
 548      c₁ * (1 - eta₁ * Real.log (↑s12 / ↑s23)) * (κ * ↑s12 / (↑s12 + ↑s23)) =
 549      c₂ * (1 - eta₂ * Real.log (↑s12 / ↑s23)) * (κ * ↑s12 / (↑s12 + ↑s23)) from by
 550        calc _ = c₁ * κ * ↑s12 / (↑s12 + ↑s23) *
 551                   (1 - eta₁ * Real.log (↑s12 / ↑s23)) := by ring
 552             _ = c₂ * κ * ↑s12 / (↑s12 + ↑s23) *
 553                   (1 - eta₂ * Real.log (↑s12 / ↑s23)) := hg23
 554             _ = _ := by ring)
 555  have hc : c₁ = c₂ := by linarith
 556  have heta : eta₁ = eta₂ := by
 557    rw [← hc] at h1
 558    have := mul_left_cancel₀ hc1 h1
 559    have : eta₁ * Real.log (↑s12 / ↑s23) = eta₂ * Real.log (↑s12 / ↑s23) := by linarith
 560    exact mul_right_cancel₀ hL this
 561  exact ⟨hc, heta⟩
 562
 563/-- For `.neg` signatures, the `cPos` field is unused and can be changed freely. -/
 564theorem neg_cPos_irrelevant (c q₁ q₂ η : ℝ) (s12 s23 : ℕ) (κ : ℝ)
 565    (hs12 : 0 < s12) (hs23 : 0 < s23) :
 566    refinedFamily ⟨c, q₁, η⟩ ⟨.neg, s12, s23, κ, hs12, hs23⟩ =
 567    refinedFamily ⟨c, q₂, η⟩ ⟨.neg, s12, s23, κ, hs12, hs23⟩ := by
 568  simp [refinedFamily]
 569
 570/-- For `.pos` signatures, the `cNeg` field is unused and can be changed freely. -/
 571theorem pos_cNeg_irrelevant (q₁ q₂ c η : ℝ) (s12 s23 : ℕ) (κ : ℝ)
 572    (hs12 : 0 < s12) (hs23 : 0 < s23) :
 573    refinedFamily ⟨q₁, c, η⟩ ⟨.pos, s12, s23, κ, hs12, hs23⟩ =
 574    refinedFamily ⟨q₂, c, η⟩ ⟨.pos, s12, s23, κ, hs12, hs23⟩ := by
 575  simp [refinedFamily]
 576
 577/-- **SECTOR CLOSURE (neg)**: For any `.neg` sector with nonzero gen12,
 578    non-degenerate cross-difference, distinct steps, and nonzero coupling,
 579    the refined family has a **unique** active-coefficient solution (c, η).
 580    This is the full `∃!` on the active parameters.
 581
 582    Combines `refinedFamily_neg_solvable` (existence) with
 583    `refinedFamily_neg_unique` (uniqueness). -/
 584theorem refined_neg_sector_closure
 585    (g12 g23 : ℝ) (s12 s23 : ℕ) (κ : ℝ)
 586    (hs12 : 0 < s12) (hs23 : 0 < s23)
 587    (hκ : κ ≠ 0) (hg12 : g12 ≠ 0)
 588    (hD : g12 * (s12 : ℝ) - g23 * (s23 : ℝ) ≠ 0)
 589    (hL : Real.log ((s12 : ℝ) / (s23 : ℝ)) ≠ 0) :
 590    ∃ c η : ℝ,
 591      refinedFamily ⟨c, 0, η⟩ ⟨.neg, s12, s23, κ, hs12, hs23⟩ = ⟨g12, g23⟩ ∧
 592      ∀ c' p' η',
 593        refinedFamily ⟨c', p', η'⟩ ⟨.neg, s12, s23, κ, hs12, hs23⟩ = ⟨g12, g23⟩ →
 594        c' = c ∧ η' = η := by
 595  obtain ⟨⟨c, cp, η⟩, hexist⟩ :=
 596    refinedFamily_neg_solvable g12 g23 s12 s23 κ hs12 hs23 hκ hg12 hD hL
 597  refine ⟨c, η, ?_, ?_⟩
 598  · -- Existence: swap cPos from cp to 0 (irrelevant for .neg)
 599    exact (neg_cPos_irrelevant c 0 cp η s12 s23 κ hs12 hs23).trans hexist
 600  · -- Uniqueness: any solution (c', p', η') must have c' = c, η' = η
 601    intro c' p' η' h'
 602    have hc_ne : c ≠ 0 := by
 603      intro hc
 604      exact hg12 (by
 605        have := congr_arg ResidualPair.gen12 hexist
 606        dsimp only [refinedFamily] at this
 607        simp [hc] at this
 608        linarith)
 609    have ⟨hc, hη⟩ := refinedFamily_neg_unique c c' cp p' η η' s12 s23 κ
 610      hs12 hs23 hκ hc_ne hL (hexist.trans h'.symm)
 611    exact ⟨hc.symm, hη.symm⟩
 612
 613/-! ## Sign-Class Family: Independent η per B_pow Sign
 614
 615Since universal η is ruled out with current data, the natural generalization
 616gives each sign class its own log-asymmetry parameter.  With 4 parameters
 617(cNeg, etaNeg, cPos, etaPos) and 4 quark equations, the system is exactly
 618determined — each sector's 2 equations uniquely fix its (c, η) pair. -/
 619
 620/-- Four global coefficients: two (amplitude, log-asymmetry) pairs,
 621    one per `B_pow` sign class. -/
 622structure SignClassCoeffs where
 623  cNeg   : ℝ
 624  etaNeg : ℝ
 625  cPos   : ℝ
 626  etaPos : ℝ
 627
 628/-- Sign-class family with independent η per sign.  Each sector's
 629    gen12/gen23 ratio is individually adjustable via its sign-class η,
 630    while the overall amplitude is set by c and coupling. -/
 631noncomputable def signClassFamily (coeffs : SignClassCoeffs)
 632    (sig : ResidualSignature) : ResidualPair :=
 633  let (c, eta) :=
 634    match sig.sign with
 635    | .neg => (coeffs.cNeg, coeffs.etaNeg)
 636    | .pos => (coeffs.cPos, coeffs.etaPos)
 637  let s12 := (sig.step12 : ℝ)
 638  let s23 := (sig.step23 : ℝ)
 639  let total := s12 + s23
 640  let logAsym := Real.log (s12 / s23)
 641  { gen12 := c * sig.coupling * s23 / total * (1 + eta * logAsym)
 642  , gen23 := -(c * sig.coupling * s12 / total * (1 - eta * logAsym))
 643  }
 644
 645/-- When both η values agree, the sign-class family collapses
 646    to the refined family with universal η. -/
 647theorem signClass_collapse_to_refined
 648    (cN cP eta : ℝ) (sig : ResidualSignature) :
 649    signClassFamily ⟨cN, eta, cP, eta⟩ sig =
 650    refinedFamily ⟨cN, cP, eta⟩ sig := by
 651  simp only [signClassFamily, refinedFamily]
 652  cases sig.sign <;> rfl
 653
 654/-! ## Concrete Instantiation: PDG Data + RS Coupling
 655
 656PDG 2022 experimental masses (MeV) at mixed reference scales.
 657Quark masses carry the standard scheme caveat: u, d, s at 2 GeV MS-bar;
 658c at m_c; b at m_b; t as pole mass.  At LO in QCD the within-sector mass
 659ratio is RG-invariant; NLO corrections are O(α_s) ≈ 12%.  Once the RG
 660bridge module compiles, quark residuals should be refined to anchor-scale
 661ratios; the present values serve as the first concrete target.
 662
 663Charged-lepton pole masses are scheme-free (no RG caveat). -/
 664
 665section ConcreteInstantiation
 666
 667def pdg_up       : ℝ := 2.16
 668def pdg_charm    : ℝ := 1270
 669def pdg_top      : ℝ := 172690
 670def pdg_down     : ℝ := 4.67
 671def pdg_strange  : ℝ := 93.4
 672def pdg_bottom   : ℝ := 4180
 673def pdg_electron : ℝ := 0.510999
 674def pdg_muon     : ℝ := 105.658
 675def pdg_tau      : ℝ := 1776.86
 676
 677/-- RS-derived strong coupling constant: α_s = 2/17 (wallpaper-group fraction).
 678    Proved to within 0.3σ of PDG α_s(M_Z) = 0.1179 ± 0.0009. -/
 679noncomputable def alphaStrong : ℝ := 2 / 17
 680
 681/-- Rung-unit residual: `log_φ(observed ratio) − integer step`.
 682    Measures the sub-leading correction in units of φ-ladder rungs. -/
 683noncomputable def rungResidual (ratio : ℝ) (step : ℕ) : ℝ :=
 684  Real.log ratio / Real.log Constants.phi - (step : ℝ)
 685
 686noncomputable def upGen12Residual : ℝ :=
 687  rungResidual (pdg_charm / pdg_up) 13
 688
 689noncomputable def upGen23Residual : ℝ :=
 690  rungResidual (pdg_top / pdg_charm) 11
 691
 692noncomputable def downGen12Residual : ℝ :=
 693  rungResidual (pdg_strange / pdg_down) 6
 694
 695noncomputable def downGen23Residual : ℝ :=
 696  rungResidual (pdg_bottom / pdg_strange) 8
 697
 698noncomputable def leptonGen12Residual : ℝ :=
 699  rungResidual (pdg_muon / pdg_electron) 11
 700
 701noncomputable def leptonGen23Residual : ℝ :=
 702  rungResidual (pdg_tau / pdg_muon) 6
 703
 704/-- Exact up-quark residual pair (from PDG mass ratios vs φ^{SDGT step}).
 705    Approximate values: gen12 ≈ +0.25, gen23 ≈ −0.79 rungs. -/
 706noncomputable def upExact : ResidualPair where
 707  gen12 := upGen12Residual
 708  gen23 := upGen23Residual
 709
 710/-- Exact down-quark residual pair (from PDG mass ratios vs φ^{SDGT step}).
 711    Approximate values: gen12 ≈ +0.23, gen23 ≈ −0.10 rungs. -/
 712noncomputable def downExact : ResidualPair where
 713  gen12 := downGen12Residual
 714  gen23 := downGen23Residual
 715
 716/-- Observed lepton residual pair (pole-mass ratios — no RG ambiguity).
 717    Approximate values: gen12 ≈ +0.08, gen23 ≈ −0.13 rungs. -/
 718noncomputable def leptonObserved : ResidualPair where
 719  gen12 := leptonGen12Residual
 720  gen23 := leptonGen23Residual
 721
 722/-! ### Specialized Closure Targets -/
 723
 724/-- **Item 8 Closure (Specialized)**: Both quark sectors fit the sign-split
 725    ratio family with coupling κ = α_s = 2/17.
 726    Proving this closes the open item and freezes the two global coefficients. -/
 727noncomputable def item8Specialized : Prop :=
 728  item8ClosureTarget upExact downExact alphaStrong alphaStrong
 729
 730/-- **All-Sector Verification**: The coefficients frozen by quark data also
 731    reproduce the observed lepton residuals — a genuine out-of-sample test.
 732
 733    Leptons share `BpowSign.neg` with up quarks, so `cNeg` is already fixed
 734    by the quark closure.  If that same `cNeg` reproduces the lepton residuals
 735    under the lepton signature `(11, 6)` and lepton coupling `κ_lep`, the
 736    family is vindicated across sign classes.
 737
 738    The parameter `kappaLepton` is left free: the leading candidate is
 739    `1 / (4 · π · 11)` (the RS electromagnetic coupling from α_seed). -/
 740noncomputable def allSectorTest (kappaLepton : ℝ) : Prop :=
 741  ∃ coeffs : RatioFamilyCoeffs,
 742    predictedResiduals coeffs (upQuarkSignature alphaStrong) = upExact ∧
 743    predictedResiduals coeffs (downQuarkSignature alphaStrong) = downExact ∧
 744    predictedResiduals coeffs (leptonSignature kappaLepton) = leptonObserved
 745
 746/-! ### Refined Closure Targets -/
 747
 748/-- **Refined Item 8 Closure**: Both quark sectors fit the log-asymmetry
 749    ratio family with κ = α_s = 2/17 and a single universal η.
 750    This is overdetermined (3 parameters, 4 equations): the `∃!` is
 751    a genuine test, not a tautology. -/
 752noncomputable def refinedItem8Specialized : Prop :=
 753  refinedItem8ClosureTarget upExact downExact alphaStrong alphaStrong
 754
 755/-- **Refined All-Sector Test**: The 3 coefficients frozen by quark data
 756    also reproduce the lepton residuals (6 equations, 3 unknowns).
 757    This is the strongest available falsification target. -/
 758noncomputable def refinedAllSectorTest (kappaLepton : ℝ) : Prop :=
 759  ∃ coeffs : RefinedCoeffs,
 760    refinedPrediction coeffs (upQuarkSignature alphaStrong) = upExact ∧
 761    refinedPrediction coeffs (downQuarkSignature alphaStrong) = downExact ∧
 762    refinedPrediction coeffs (leptonSignature kappaLepton) = leptonObserved
 763
 764/-! ### Per-Sector η Values (from closed-form formula)
 765
 766Each sector's η is computed via `etaFromData` using PDG residuals and
 767SDGT steps.  The lepton value is small and perturbative (|η·L| ≈ 4%),
 768confirming the family form works for scheme-free pole masses.  Quark
 769values are inflated by mixed-scale PDG artifacts. -/
 770
 771/-- Lepton η from pole-mass residuals: +0.065.
 772    This is the cleanest data point — no RG ambiguity. -/
 773noncomputable def leptonEta : ℝ :=
 774  etaFromData leptonGen12Residual leptonGen23Residual 11 6
 775
 776/-- Up-quark η from mixed-scheme PDG: −2.72.
 777    Inflated by the t/c mixed-scheme comparison (pole vs MS-bar at m_c). -/
 778noncomputable def upQuarkEta : ℝ :=
 779  etaFromData upGen12Residual upGen23Residual 13 11
 780
 781/-- Down-quark η from mixed-scheme PDG: −0.88. -/
 782noncomputable def downQuarkEta : ℝ :=
 783  etaFromData downGen12Residual downGen23Residual 6 8
 784
 785/-! ### Concrete Lepton Closure
 786
 787The lepton sector is the cleanest `.neg` test case because the masses are pole
 788masses (no QCD scheme ambiguity).  The non-degeneracy hypotheses needed by
 789`refined_neg_sector_closure` can be verified directly from φ-power comparisons:
 790
 791* `μ/e > φ¹¹`, so `leptonGen12Residual > 0`
 792* `τ/μ < φ⁶`, so `leptonGen23Residual < 0`
 793
 794This gives the full `∃!` closure for the concrete lepton signature `(11, 6)`
 795at the candidate electromagnetic coupling `κ_lep = 1 / alpha_seed`. -/
 796
 797/-- Leading lepton coupling candidate from the RS electromagnetic seed:
 798    `κ_lep = 1 / alpha_seed = 1 / (4π·11)`. -/
 799noncomputable def kappaLeptonCandidate : ℝ := 1 / Constants.alpha_seed
 800
 801theorem kappaLeptonCandidate_pos : 0 < kappaLeptonCandidate := by
 802  unfold kappaLeptonCandidate Constants.alpha_seed
 803  positivity
 804
 805theorem kappaLeptonCandidate_ne_zero : kappaLeptonCandidate ≠ 0 :=
 806  ne_of_gt kappaLeptonCandidate_pos
 807
 808/-- The lepton step log-asymmetry is positive: `log(11/6) > 0`. -/
 809theorem leptonLogAsym_pos : 0 < Real.log ((11 : ℝ) / 6) := by
 810  have hratio_gt_one : (1 : ℝ) < (11 : ℝ) / 6 := by norm_num
 811  exact Real.log_pos hratio_gt_one
 812
 813theorem leptonLogAsym_ne_zero : Real.log ((11 : ℝ) / 6) ≠ 0 :=
 814  ne_of_gt leptonLogAsym_pos
 815
 816/-- The μ/e residual is positive because the observed ratio exceeds `φ¹¹`. -/
 817theorem leptonGen12Residual_pos : 0 < leptonGen12Residual := by
 818  unfold leptonGen12Residual rungResidual
 819  have hlogphi : 0 < Real.log Constants.phi := Real.log_pos Constants.one_lt_phi
 820  have hphi11_lt : Constants.phi ^ (11 : ℕ) < pdg_muon / pdg_electron := by
 821    have hphi11_lt_200 : Constants.phi ^ (11 : ℕ) < (200 : ℝ) := by
 822      rw [Constants.phi_eleventh_eq]
 823      linarith [Constants.phi_lt_onePointSixTwo]
 824    have hratio_gt_200 : (200 : ℝ) < pdg_muon / pdg_electron := by
 825      norm_num [pdg_muon, pdg_electron]
 826    linarith
 827  have hmain : (11 : ℝ) < Real.log (pdg_muon / pdg_electron) / Real.log Constants.phi := by
 828    apply (lt_div_iff₀ hlogphi).2
 829    rw [show (11 : ℝ) = (11 : ℕ) by norm_num, ← Real.log_rpow Constants.phi_pos]
 830    apply Real.log_lt_log
 831    · exact Real.rpow_pos_of_pos Constants.phi_pos 11
 832    · simpa [Real.rpow_natCast] using hphi11_lt
 833  linarith
 834
 835theorem leptonGen12Residual_ne_zero : leptonGen12Residual ≠ 0 :=
 836  ne_of_gt leptonGen12Residual_pos
 837
 838/-- The τ/μ residual is negative because the observed ratio lies below `φ⁶`. -/
 839theorem leptonGen23Residual_neg : leptonGen23Residual < 0 := by
 840  unfold leptonGen23Residual rungResidual
 841  have hlogphi : 0 < Real.log Constants.phi := Real.log_pos Constants.one_lt_phi
 842  have hratio_lt : pdg_tau / pdg_muon < Constants.phi ^ (6 : ℕ) := by
 843    have hratio_lt_17 : pdg_tau / pdg_muon < (17 : ℝ) := by
 844      norm_num [pdg_tau, pdg_muon]
 845    have h17_lt_phi6 : (17 : ℝ) < Constants.phi ^ (6 : ℕ) := by
 846      rw [Constants.phi_sixth_eq]
 847      linarith [Constants.phi_gt_onePointFive]
 848    linarith
 849  have hratio_pos : 0 < pdg_tau / pdg_muon := by
 850    norm_num [pdg_tau, pdg_muon]
 851  have hmain : Real.log (pdg_tau / pdg_muon) / Real.log Constants.phi < (6 : ℝ) := by
 852    apply (div_lt_iff₀ hlogphi).2
 853    rw [show (6 : ℝ) = (6 : ℕ) by norm_num, ← Real.log_rpow Constants.phi_pos]
 854    apply Real.log_lt_log
 855    · exact hratio_pos
 856    · simpa [Real.rpow_natCast] using hratio_lt
 857  linarith
 858
 859/-- The lepton cross-difference is strictly positive, hence nonzero. -/
 860theorem leptonCrossDiff_pos :
 861    0 < leptonGen12Residual * (11 : ℝ) - leptonGen23Residual * (6 : ℝ) := by
 862  nlinarith [leptonGen12Residual_pos, leptonGen23Residual_neg]
 863
 864theorem leptonCrossDiff_ne_zero :
 865    leptonGen12Residual * (11 : ℝ) - leptonGen23Residual * (6 : ℝ) ≠ 0 :=
 866  ne_of_gt leptonCrossDiff_pos
 867
 868/-- Concrete `∃!` closure for the lepton `.neg` sector at the candidate
 869    electromagnetic coupling `κ_lep = 1 / (4π·11)`.  This freezes the active
 870    pair `(cNeg, eta)` independently of the quark data. -/
 871theorem leptonSectorClosure :
 872    ∃ c η : ℝ,
 873      refinedFamily ⟨c, 0, η⟩ (leptonSignature kappaLeptonCandidate) = leptonObserved ∧
 874      ∀ c' p' η',
 875        refinedFamily ⟨c', p', η'⟩ (leptonSignature kappaLeptonCandidate) = leptonObserved →
 876        c' = c ∧ η' = η := by
 877  have hs12 : 0 < 11 := by norm_num
 878  have hs23 : 0 < 6 := by norm_num
 879  simpa [leptonSignature, leptonObserved] using
 880    refined_neg_sector_closure leptonGen12Residual leptonGen23Residual 11 6
 881      kappaLeptonCandidate hs12 hs23 kappaLeptonCandidate_ne_zero
 882      leptonGen12Residual_ne_zero leptonCrossDiff_ne_zero leptonLogAsym_ne_zero
 883
 884/-- Closed-form lepton anchor for the active `.neg` amplitude.  This is the
 885    explicit `cNeg` obtained from the lepton gen12 equation once `eta` is fixed
 886    to `leptonEta`. -/
 887noncomputable def leptonAnchoredCNeg (kappaLepton : ℝ) : ℝ :=
 888  leptonGen12Residual * ((11 : ℝ) + 6) /
 889    (kappaLepton * 6 * (1 + leptonEta * Real.log ((11 : ℝ) / 6)))
 890
 891/-- The lepton-anchored global coefficient package.  `cPos` is left at `0`
 892    because leptons only freeze the `.neg` branch. -/
 893noncomputable def leptonAnchoredCoeffs (kappaLepton : ℝ) : RefinedCoeffs :=
 894  ⟨leptonAnchoredCNeg kappaLepton, 0, leptonEta⟩
 895
 896/-! ### Anchor-Scale Quark Transport (LO scaffold)
 897
 898The current LO scaffold transports the PDG quark masses to the RS anchor scale
 899`μ* = 182.201 GeV` using piecewise one-loop α_s and the one-loop mass anomalous
 900dimension.  This is the concrete bridge needed to compare lepton-frozen
 901predictions against scheme-consistent quark residuals. -/
 902
 903/-- One-loop α_s in the `n_f = 6` region, anchored at `μ*`. -/
 904noncomputable def alphaS6At (μ : ℝ) : ℝ :=
 905  Physics.RG.alpha_s_running Physics.RG.rs_alpha_s_anchor (Physics.RG.b0_qcd 6)
 906    μ Physics.RG.rs_anchor_scale
 907
 908/-- Boundary value α_s(m_t) obtained from the `n_f = 6` branch. -/
 909noncomputable def alphaSAtTopThreshold : ℝ :=
 910  alphaS6At Physics.RG.top_threshold.scale
 911
 912/-- One-loop α_s in the `n_f = 5` region, matched at the top threshold. -/
 913noncomputable def alphaS5At (μ : ℝ) : ℝ :=
 914  Physics.RG.alpha_s_running alphaSAtTopThreshold (Physics.RG.b0_qcd 5)
 915    μ Physics.RG.top_threshold.scale
 916
 917/-- Boundary value α_s(m_b) obtained from the `n_f = 5` branch. -/
 918noncomputable def alphaSAtBottomThreshold : ℝ :=
 919  alphaS5At Physics.RG.bottom_threshold.scale
 920
 921/-- One-loop α_s in the `n_f = 4` region, matched at the bottom threshold. -/
 922noncomputable def alphaS4At (μ : ℝ) : ℝ :=
 923  Physics.RG.alpha_s_running alphaSAtBottomThreshold (Physics.RG.b0_qcd 4)
 924    μ Physics.RG.bottom_threshold.scale
 925
 926/-- Boundary value α_s(m_c) obtained from the `n_f = 4` branch. -/
 927noncomputable def alphaSAtCharmThreshold : ℝ :=
 928  alphaS4At Physics.RG.charm_threshold.scale
 929
 930/-- One-loop α_s in the `n_f = 3` region, matched at the charm threshold. -/
 931noncomputable def alphaS3At (μ : ℝ) : ℝ :=
 932  Physics.RG.alpha_s_running alphaSAtCharmThreshold (Physics.RG.b0_qcd 3)
 933    μ Physics.RG.charm_threshold.scale
 934
 935/-- Piecewise one-loop α_s stitched across the heavy-quark thresholds. -/
 936noncomputable def alphaSPiecewise (μ : ℝ) : ℝ :=
 937  if Physics.RG.top_threshold.scale ≤ μ then alphaS6At μ
 938  else if Physics.RG.bottom_threshold.scale ≤ μ then alphaS5At μ
 939  else if Physics.RG.charm_threshold.scale ≤ μ then alphaS4At μ
 940  else alphaS3At μ
 941
 942/-- Shared PDG reference scale for the light MS-bar quarks (`u,d,s`). -/
 943def quarkReferenceScale2GeV : ℝ := 2
 944
 945/-- Threshold lists for upward transport to the anchor scale. -/
 946def thresholdsFromTwoGeV : List Physics.RG.FlavorThreshold :=
 947  [Physics.RG.bottom_threshold, Physics.RG.top_threshold]
 948
 949def thresholdsFromCharm : List Physics.RG.FlavorThreshold :=
 950  [Physics.RG.bottom_threshold, Physics.RG.top_threshold]
 951
 952def thresholdsFromBottom : List Physics.RG.FlavorThreshold :=
 953  [Physics.RG.top_threshold]
 954
 955/-- Light and heavy quark masses transported to the anchor scale `μ*`. -/
 956noncomputable def upMassAtAnchor : ℝ :=
 957  Physics.RG.transport_mass_through pdg_up alphaSPiecewise quarkReferenceScale2GeV
 958    Physics.RG.rs_anchor_scale thresholdsFromTwoGeV 4
 959
 960noncomputable def downMassAtAnchor : ℝ :=
 961  Physics.RG.transport_mass_through pdg_down alphaSPiecewise quarkReferenceScale2GeV
 962    Physics.RG.rs_anchor_scale thresholdsFromTwoGeV 4
 963
 964noncomputable def strangeMassAtAnchor : ℝ :=
 965  Physics.RG.transport_mass_through pdg_strange alphaSPiecewise quarkReferenceScale2GeV
 966    Physics.RG.rs_anchor_scale thresholdsFromTwoGeV 4
 967
 968noncomputable def charmMassAtAnchor : ℝ :=
 969  Physics.RG.transport_mass_through pdg_charm alphaSPiecewise
 970    Physics.RG.charm_threshold.scale Physics.RG.rs_anchor_scale thresholdsFromCharm 4
 971
 972noncomputable def bottomMassAtAnchor : ℝ :=
 973  Physics.RG.transport_mass_through pdg_bottom alphaSPiecewise
 974    Physics.RG.bottom_threshold.scale Physics.RG.rs_anchor_scale thresholdsFromBottom 5
 975
 976noncomputable def topMassAtAnchor : ℝ :=
 977  Physics.RG.transport_mass_through pdg_top alphaSPiecewise
 978    Physics.RG.top_threshold.scale Physics.RG.rs_anchor_scale [] 6
 979
 980/-- Anchor-scale quark residuals built from the transported masses.  Under the
 981    current LO scaffold, shell evaluation gives approximately:
 982    up `(0.02, +0.57)` and down `(0.23, +0.19)` in rung units. -/
 983noncomputable def anchorUpGen12Residual : ℝ :=
 984  rungResidual (charmMassAtAnchor / upMassAtAnchor) 13
 985
 986noncomputable def anchorUpGen23Residual : ℝ :=
 987  rungResidual (topMassAtAnchor / charmMassAtAnchor) 11
 988
 989noncomputable def anchorDownGen12Residual : ℝ :=
 990  rungResidual (strangeMassAtAnchor / downMassAtAnchor) 6
 991
 992noncomputable def anchorDownGen23Residual : ℝ :=
 993  rungResidual (bottomMassAtAnchor / strangeMassAtAnchor) 8
 994
 995noncomputable def anchorUpExact : ResidualPair where
 996  gen12 := anchorUpGen12Residual
 997  gen23 := anchorUpGen23Residual
 998
 999noncomputable def anchorDownExact : ResidualPair where
1000  gen12 := anchorDownGen12Residual
1001  gen23 := anchorDownGen23Residual
1002
1003/-- Up-sector out-of-sample prediction after freezing `(cNeg, eta)` on leptons. -/
1004noncomputable def leptonAnchoredUpPrediction (kappaLepton : ℝ) : ResidualPair :=
1005  refinedPrediction (leptonAnchoredCoeffs kappaLepton) (upQuarkSignature alphaStrong)
1006
1007/-- Two down-sector amplitudes implied by the transported anchor residuals once
1008    the lepton value of `eta` is frozen.  Equality of these two expressions is
1009    the one-parameter down-sector consistency test under lepton anchoring. -/
1010noncomputable def downAnchorCPosFromGen12 : ℝ :=
1011  anchorDownGen12Residual * ((6 : ℝ) + 8) /
1012    (alphaStrong * 8 * (1 + leptonEta * Real.log ((6 : ℝ) / 8)))
1013
1014noncomputable def downAnchorCPosFromGen23 : ℝ :=
1015  -(anchorDownGen23Residual * ((6 : ℝ) + 8) /
1016    (alphaStrong * 6 * (1 - leptonEta * Real.log ((6 : ℝ) / 8))))
1017
1018/-- The concrete anchor-scale lepton-anchored falsification target:
1019    the up prediction must match the transported anchor residuals, and the down
1020    sector must induce a single consistent `cPos`. -/
1021noncomputable def leptonAnchoredAnchorTest : Prop :=
1022  leptonAnchoredUpPrediction kappaLeptonCandidate = anchorUpExact ∧
1023  downAnchorCPosFromGen12 = downAnchorCPosFromGen23
1024
1025/-! ### Lepton-Anchored Closure Strategy
1026
1027Since leptons provide scheme-free data with perturbative η ≈ 0.065,
1028the resolution strategy is:
1029
10301. **Anchor on leptons**: Fix η and cNeg from the two lepton equations
1031   (both generations share `BpowSign.neg` with up quarks).
10322. **Predict quarks at anchor scale**: The frozen (cNeg, η) pair
1033   predicts up-quark residuals at the anchor scale μ* = 182 GeV.
1034   Down quarks provide an independent prediction via cPos.
10353. **RG bridge**: Compile the `RunningCouplings` module to transport
1036   PDG quark masses to the anchor scale, producing scheme-consistent
1037   residuals for comparison.
1038
1039This makes the quark mass predictions genuine out-of-sample tests
1040rather than fits, and keeps the family form validated by clean data. -/
1041
1042/-- Lepton-anchored closure: the refined family with lepton-derived η
1043    fits both leptons (by construction) and predicts anchor-scale quarks.
1044    The proposition is: there exist global coefficients such that
1045    the family simultaneously fits leptons at coupling κ_lep and
1046    predicts quark residuals at the anchor scale. -/
1047noncomputable def leptonAnchoredTarget
1048    (kappaLepton : ℝ)
1049    (anchorUpExact anchorDownExact : ResidualPair) : Prop :=
1050  ∃ coeffs : RefinedCoeffs,
1051    refinedPrediction coeffs (leptonSignature kappaLepton) = leptonObserved ∧
1052    refinedPrediction coeffs (upQuarkSignature alphaStrong) = anchorUpExact ∧
1053    refinedPrediction coeffs (downQuarkSignature alphaStrong) = anchorDownExact
1054
1055/-- Sign-class family closure: the 4-parameter family with independent η
1056    per sign class fits all three sectors simultaneously.
1057    This is the full all-sector test (6 equations, 4 unknowns + 1 free κ_lep). -/
1058noncomputable def signClassAllSectorTarget (kappaLepton : ℝ) : Prop :=
1059  ∃ coeffs : SignClassCoeffs,
1060    signClassFamily coeffs (upQuarkSignature alphaStrong) = upExact ∧
1061    signClassFamily coeffs (downQuarkSignature alphaStrong) = downExact ∧
1062    signClassFamily coeffs (leptonSignature kappaLepton) = leptonObserved
1063
1064end ConcreteInstantiation
1065
1066end Item8ClosureTarget
1067end Verification
1068end IndisputableMonolith
1069

source mirrored from github.com/jonwashburn/shape-of-logic