Skip to main content
Falsification surface · 2026

How to kill TFPT.

A theory that cannot fail explains nothing. This page lists the committed kill criteria of the TFPT 5.4 compiler closure — frozen in advance in the freeze file — together with the experiment or structural argument that would kill the construction. The no-knobs audit underneath records what the theory may consume as input and what it may not.

Each output is killed by any single satisfied criterion. The construction does not survive on most outputs being right.

The kill board

Every readout, its kill condition, and where it stands now

The full status-graded surface as a board: each card carries the predicted value, the single condition that would falsify it, and — where a standalone empirical audit exists in the experiments/ tree — its live status. These confrontations are search targets, not load-bearing claims; no card is upgraded by data proximity.

Status legendLive · consistentLive · robustLive · tensionLive · nullData-limitedStructural / committed

  • α⁻¹(0)

    [E]
    α⁻¹(0) = 137.035 999 216 8…
    Kill condition

    Failure of the unique-root equation F_U(1)(α) = 0, a second admissible root, or a stable mismatch outside the stated interface uncertainty.

    Structural / committedSource document

  • λ_C

    [E]
    λ_C = 0.22438
    Kill condition

    Stable CKM global-fit mismatch after the declared comparison map.

    Structural / committedSource document

  • det R, χ_R

    [E]
    det R = 8, minors (2,3,5), χ_R = t³ − 9t² + 10t − 8
    Kill condition

    A future CKM/PMNS global fit that cannot be carried by a residue matrix with det 8, principal minors (2,3,5) and this characteristic polynomial.

    Structural / committedSource document

  • Q_Koide

    [C]
    Q = 0.664 (target 2/3 = |ℤ₂|/N_fam)
    Kill condition

    A source→pole transfer that lands far from 2/3, or a demonstration that the lepton φ₀-ladder is incompatible with the measured charged-lepton masses.

    Structural / committedSource document

  • sin²θ₁₂

    [E]
    sin²θ₁₂ = 1/3 − φ₀/2 = 0.3067
    Kill condition

    A JUNO central value clearly away from 0.307 at high significance kills the seam-misalignment mechanism.

    Live · consistentEmpirical audit

  • sin²θ₁₃

    [E]
    sin²θ₁₃ = φ₀ e^(−5/6) = 0.0231
    Kill condition

    Robust normal-ordering global-fit exclusion at the stated confidence level.

    Live · consistentEmpirical audit

  • sin²θ₂₃

    [C]
    sin²θ₂₃ ≈ 1/2 (octant not selected)
    Kill condition

    A robust off-maximal octant determination by NOvA / T2K / DUNE.

    Structural / committedSource document

  • NO, m_ββ

    [C]
    Normal ordering, small m_ββ
    Kill condition

    Inverted ordering, a large m_ββ detection, a cosmological Σm_ν < 0.0586 eV (the normal-ordering floor, v272), or a δ_PMNS robustly outside the band 240° ± ~9° = δ_CKM,lead + π (the Galois lock; v320, sharpened v322) kills the minimal Majorana / Galois-CP branch.

    Structural / committedSource document

  • δ_PMNS

    [C]
    δ_PMNS = 240° (band 240° ± ~9°)
    Kill condition

    A measured δ_PMNS robustly outside the band 240° ± ~9° (= δ_CKM,lead + π), or landing on a different hexagonal node (60/120/180/300°), at >3σ (DUNE / Hyper-K / JUNO) falsifies the whole Galois-CP organisation.

    Live · consistentEmpirical audit

  • θ_eff = 0

    [E]
    θ_eff = 0 (structural null)
    Kill condition

    A solid neutron-EDM signal above the SM background falsifies the structural cancellation.

    Live · consistentEmpirical audit

  • m_p/m_e

    [O]
    m_p/m_e = 1836.15 (explicitly not claimed)
    Kill condition

    Only fails if mis-asserted as a compiler power; there is no clean φ₀ power for the QCD-confinement / EW-Yukawa ratio.

    Structural / committedSource document

  • n_s

    [C]
    n_s = 1 − 2/N★ ∈ [0.960, 0.967] (frozen band)
    Kill condition

    A robust n_s far from the Starobinsky line on the same R² attractor (n_s ≥ 0.967 also kills the scalaron-reheating chain).

    Live · consistentEmpirical audit

  • r

    [C]
    r = 12/N★² ∈ [0.0033, 0.0048] (frozen band)
    Kill condition

    Any robust r ≳ 0.01 kills the R² branch carrying M_Pl and A_s.

    Data-limitedEmpirical audit

  • A_s

    [C]
    A_s = N★² c₃⁷/(24π²) ≈ 2.0×10⁻⁹
    Kill condition

    A_s incompatible with the seam-fixed scalaron mass on the R² branch.

    Live · tensionEmpirical audit

  • α_s

    [C]
    α_s = −2/N★² = −r/6 ∈ [−8.0, −5.6]×10⁻⁴ (LO band)
    Kill condition

    Robust α_s < −5×10⁻³ or > +3×10⁻³ at ≥5σ; any robust POSITIVE running kills the plateau branch (plateau potentials give α_s < 0).

    Live · consistentEmpirical audit

  • μ

    [C]
    μ ∈ [1.5, 2.3]×10⁻⁸ (band); 1.6×10⁻⁸ sharp / 2.0×10⁻⁸ profiled
    Kill condition

    Robust μ < 0.9×10⁻⁸ or > 4×10⁻⁸.

    Data-limitedEmpirical audit

  • M_scal

    [E]
    M = c₃^(7/2) M̄ = 3.06×10¹³ GeV
    Kill condition

    A scalaron mass incompatible with the seam power c₃⁷ = c₃^(Ω_adm − 10 b₁).

    Structural / committedSource document

  • Ω_b

    [C]
    Ω_b = (4π − 1)β_rad = 0.04894
    Kill condition

    Robust inconsistency under the declared Planck comparison convention.

    Live · consistentEmpirical audit

  • η_B

    [C]
    η_B = 6.1×10⁻¹⁰
    Kill condition

    Robust exclusion of the quoted value under the declared cosmological pipeline (as a compiler power it is explicitly not closed).

    Live · consistentEmpirical audit

  • H₀ ∼ √Λ

    [C]
    v_EW ∼ e^(−α⁻¹/5), Λ ∼ e^(−2α⁻¹), H₀ ∼ √Λ
    Kill condition

    A robust w ≠ −1 kills the single-engine dark-energy readout.

    Data-limitedEmpirical audit

  • N_Φ = 1

    [E]
    exactly one seam-even light doublet
    Kill condition

    Robust discovery of a second light seam-even Higgs doublet.

    Structural / committedSource document

  • λ(M̄) ≈ 0

    [C]
    λ(M̄_Pl) = 0 and β_λ(M̄_Pl) = 0 (double criticality); m_H band 129–134 GeV
    Kill condition

    A settled (m_t, α_s) RGE pull off the double-critical surface λ(M̄) = β_λ(M̄) = 0 at >5σ kills the free-seam boundary condition (the compiler core is untouched).

    Live · consistentEmpirical audit

  • β_rad

    [E]
    β_rad = φ₀/(4π) = 0.2424°
    Kill condition

    An externally calibrated β = 0 within tight error.

    Live · consistentEmpirical audit

  • m_a

    [C]
    m_a ≈ 23.8 µeV, f_a = M_scal/128, g_aγγ coefficient −4c₃
    Kill condition

    Exclusion of the determinant-line axion window at the coupled sensitivity (relic abundance is scenario-sensitive, not closed).

    Live · robustEmpirical audit

  • BR(K→πνν̄)

    [C]
    BR(K⁺) = 9.45×10⁻¹¹, BR(K_L) = 3.33×10⁻¹¹
    Kill condition

    A stable NA62 BR(K⁺) outside [7,12]×10⁻¹¹, or a KOTO-II BR(K_L) off the predicted Grossman–Nir point, breaks the TFPT flavor bridge for this sector (the compiler core is untouched).

    Live · consistentEmpirical audit

  • β_BH

    [C]
    β_BH ∝ 16c₃⁴ · Q_e Q_m / r² (achromatic, 1/r², sign-flipping)
    Kill condition

    A structured achromatic residual intercept consistent with zero across the horizon-scale image after honest GRMHD subtraction — or one failing the frequency / 1/r² / sign-flip nulls — falsifies the channel (the compiler core is untouched).

    Data-limitedEmpirical audit

  • ν₃ = (3/2)ν_u

    [O]
    third tooth at ν₃ = 1.5 ν_u (661.5 / 414 / 252 / 363 Hz); integer harmonics forbidden
    Kill condition

    A detected integer line (e.g. 4ν₀ = 2ν_u/1.5) with no geometric tooth at 1.5ν_u kills the ladder reading of the 3:2 pairs; even a tooth hit would stay [C] until the ladder↔oscillator mapping is derived. The compiler core is untouched either way.

    Live · nullEmpirical audit
The confrontation table

Every readout against the data, with its decisive year

The same prediction surface as one dense table: for each readout, the derivation, the TFPT value, the current measured value, the deviation, the dated experimental source, and the experiment — with an approximate year — that will turn it into a definitive hit or kill (typically a sharper, future measurement). Every value is repo-documented; nothing is fitted.

Confrontation table — value · derivation · TFPT · measured · deviation · source · decisive year27 readouts
Every TFPT prediction-surface readout with its derivation, the predicted value, the current measured value, the deviation, the dated experimental source, and the experiment and approximate year that will make it a definitive hit or kill.
ReadoutDerivationTFPT valueMeasured / experimentalDeviationSource & dateDecisive — hit or kill
α⁻¹(0)[E]
Source document
FU(1)(α)=0, ∑L+NΦ=41=10 b1F_{U(1)}(\alpha)=0,\ \textstyle\sum L + N_\Phi = 41 = 10\,b_1
137.035 999 217137.035 999 177 ± 2.1×10⁻⁸+1.9σCODATA 2022CODATA / atom-recoil refinement (ongoing)
λ_C[E]
Source document
λC=φ0(1−φ0)\lambda_C = \sqrt{\varphi_0(1-\varphi_0)}
0.224380.2245 ± 0.0005−0.24σPDG 2024 (|V_us|)CKM global fit (ongoing)
det R, χ_R[E]
Source document
det⁡R=h(D5)=8, minors=(2,3,5)\det R = h(D_5) = 8,\ \mathrm{minors}=(2,3,5)
det R = 8, minors (2,3,5)carried by current CKM/PMNS global fitsstructural (no fit avoids it)CKM/PMNS global fits (PDG / NuFIT 2024)future CKM/PMNS global fit
Q_Koide[C]
Source document
Q⋆=∣Z2∣Nfam=23Q_\star = \tfrac{|\mathbb{Z}_2|}{N_{\mathrm{fam}}} = \tfrac{2}{3}
0.664 (target 2/3 = 0.6667)Q_exp ≈ 0.6666−0.33% (near-miss, [C])PDG charged-lepton pole masses 2024structural — no single experiment (source→pole transfer)
sin²θ₁₂[E]
Empirical audit
sin⁡2θ12=13−φ02\sin^2\theta_{12} = \tfrac{1}{3} - \tfrac{\varphi_0}{2}
0.306750.307 ± 0.012 (NuFIT) · 0.3092 ± 0.0087 (JUNO)−0.02σ (NuFIT) · −0.28σ (JUNO)NuFIT 6.0 (2024) · JUNO (2025)JUNO ~2026–2028 (the fastest falsifier)
sin²θ₁₃[E]
Empirical audit
sin⁡2θ13=φ0 e−5/6\sin^2\theta_{13} = \varphi_0\,e^{-5/6}
0.02310.02195 ± 0.00058+2.0σ (largest single pull; θ₂₃/V_cb sit at ~1.8σ — one named post-hoc [O] pattern candidate, v467, record unchanged)NuFIT 6.0 (2024)JUNO / global fit ~2026–2028
sin²θ₂₃[C]
Source document
θ23=45∘ (μτ-symmetric)\theta_{23} = 45^\circ\ (\mu\tau\text{-symmetric})
≈ 0.5 (octant not selected)0.470 ± 0.017+1.76σ (octant open; v467 [O] candidate cos 2θ₂₃ = φ₀ would give +0.20σ)NuFIT 6.0 (2024)DUNE / NOvA / T2K ~2030 (octant)
NO, m_ββ[C]
Source document
normal ordering, Σmν≳0.0586 eV\text{normal ordering},\ \Sigma m_\nu \gtrsim 0.0586\,\mathrm{eV}
NO, small m_ββ, Σm_ν ≳ 0.0586 eVNO preferred; Σm_ν < ~0.072 eVconsistent (NO floor at the cosmological edge)NuFIT 6.0 (2024) · DESI+Planck (2024)DUNE / LEGEND / nEXO ~2028–2035
δ_PMNS[C]
Empirical audit
δPMNS=arg⁡(ζ64)=4 δCKMlead\delta_{\mathrm{PMNS}} = \arg(\zeta_6^4) = 4\,\delta_{\mathrm{CKM}}^{\mathrm{lead}}
240° (band 240° ± ~9°)212°⁺²⁶₋₄₁ (NO best fit)+1.08σ (consistent, weak power)NuFIT 6.0 (2024)DUNE / Hyper-K ~2030+
θ_eff = 0[E]
Empirical audit
arg⁡det⁡Mu=arg⁡det⁡Md=0⇒θeff=0\arg\det M_u = \arg\det M_d = 0 \Rightarrow \theta_{\mathrm{eff}} = 0
0 (structural null)|d_n| < 1.8×10⁻²⁶ e·cm (consistent with 0)null (no EDM signal)PSI nEDM (2020)PSI n2EDM / SNS (ongoing)
m_p/m_e[O]
Source document
mpme=ΛQCDEW Yukawa (not a power)\tfrac{m_p}{m_e} = \tfrac{\Lambda_{\mathrm{QCD}}}{\text{EW Yukawa}}\ (\text{not a power})
not claimed (≈ 1836)1836.152673n/a (explicitly not claimed)CODATA 2022structural — fails only if mis-asserted
n_s[C]
Empirical audit
ns=1−2N⋆, N⋆∈[50,60]n_s = 1 - \tfrac{2}{N_\star},\ N_\star \in [50,60]
0.960–0.967 (0.9611 at N★=51.4)0.9649 ± 0.0042−0.91σ (band consistent)Planck 2018CMB-S4 ~2028–2032
r[C]
Empirical audit
r=12N⋆2, N⋆∈[50,60]r = \tfrac{12}{N_\star^2},\ N_\star \in [50,60]
0.0033–0.0048 (0.0045)< 0.036 (95% upper limit)below bound (data-limited)BICEP/Keck BK18 (2021)CMB-S4 / LiteBIRD ~2028–2032 (σ_r ~ 5×10⁻⁴)
A_s[C]
Empirical audit
As=N⋆224π2 c37A_s = \tfrac{N_\star^2}{24\pi^2}\,c_3^7
≈ 2.0×10⁻⁹≈ 2.1×10⁻⁹consistent in band (N★ ≈ 56 profiled)Planck 2018fixed by N★ / reheating speed
α_s[C]
Empirical audit
αs=−2N⋆2, N⋆∈[50,60]\alpha_s = -\tfrac{2}{N_\star^2},\ N_\star \in [50,60]
−7.1×10⁻⁴ (band −8.0…−5.6×10⁻⁴)−0.0045 ± 0.0067 (Planck) · +0.0062 ± 0.0052 (P-ACT-LB)+0.57σ (Planck) · −1.33σ (P-ACT-LB record leg)Planck 2018 X · ACT DR6 (Calabrese+ 2025)CMB-S4 + DESI + Euclid (σ(α_s) ~ 3.3×10⁻⁴, systematics-limited)
μ[C]
Empirical audit
μ≈2.2∫PζWμ dln⁡k\mu \approx 2.2 \int P_\zeta W_\mu \, d\ln k
1.5–2.3×10⁻⁸ (1.6×10⁻⁸ sharp / 2.0×10⁻⁸ profiled)|μ| < 9×10⁻⁵ (95%)data-limited (4 dex below the bound)COBE/FIRAS (1996)PIXIE-class / Voyage-2050 (σ_μ ~ 10⁻⁹): 3σ branch decision
M_scal[E]
Source document
Mscal2MˉPl2=c3 7\tfrac{M_{\mathrm{scal}}^2}{\bar M_{\mathrm{Pl}}^2} = c_3^{\,7}
3.06×10¹³ GeVcanonical Starobinsky value (from A_s)matches the A_s-inferred massPlanck 2018 (A_s)CMB-S4 inflation constraints ~2028–2032
Ω_b[C]
Empirical audit
Ωb=(4π−1) βrad\Omega_b = (4\pi - 1)\,\beta_{\mathrm{rad}}
0.048940.0493 ± 0.0006 (Planck) · 0.0483 ± 0.0072 (FRB)0.04σ (BBN leg) · 0.10σ (FRB)Planck 2018 · FRB Macquart DM(z)settled — consistency, weak discriminator
η_B[C]
Empirical audit
ηB=273.9×10−10 Ωbh2\eta_B = 273.9\times10^{-10}\,\Omega_b h^2
6.09×10⁻¹⁰ (Boltzmann solve 6.5×10⁻¹⁰)6.1×10⁻¹⁰×1.07 (consistent, [C] not closed)Planck 2018 / BBNsettled observation — not a discriminator
H₀ ∼ √Λ[C]
Empirical audit
Λ∼e−2α−1, H0∼Λ, w=−1\Lambda \sim e^{-2\alpha^{-1}},\ H_0 \sim \sqrt{\Lambda},\ w = -1
w = −1 (single engine)DESI DR2 hints w ≠ −1w = −1 excluded at 4.4σ (1 combo; watchdog ARMED)DESI DR2 (2025)DESI / Euclid (ongoing)
N_Φ = 1[E]
Source document
NΦ=gcar−∣μ4∣=1N_\Phi = g_{\mathrm{car}} - |\mu_4| = 1
exactly 1 light doublet1 Higgs doublet observed (no 2nd)consistent (structural prohibition)LHC (ATLAS / CMS) 2024HL-LHC (ongoing)
λ(M̄) ≈ 0[C]
Empirical audit
λ(MˉPl)=0=βλ (free seam)\lambda(\bar M_{\mathrm{Pl}}) = 0 = \beta_\lambda \ \text{(free seam)}
m_H = 133.5 GeV (band 129–134)125.25 ± 0.17 GeV (λ(M_Pl) = −0.0143 ± 0.0057)a few GeV below the boundary (metastable, 2.5σ)PDG 2024 · Buttazzo+ 2013 NNLO fitprecision m_t / α_s (FCC-ee class) sharpen the surface
β_rad[E]
Empirical audit
βrad=φ04π\beta_{\mathrm{rad}} = \tfrac{\varphi_0}{4\pi}
0.2424°0.215° ± 0.074°+0.37σACT DR6 (2025)LiteBIRD / Simons Obs. ~2028+
m_a[C]
Empirical audit
fa=Mscal128, ma≈23.8 μeVf_a = \tfrac{M_{\mathrm{scal}}}{128},\ m_a \approx 23.8\,\mu\mathrm{eV}
23.8 µeV (f_a = M_scal/128)not excluded (HAYSTAC band)data-limited (DFSZ/KSVZ not yet reached)haloscopes (HAYSTAC 2024)haloscope coverage at 23.8 µeV (this decade)
BR(K→πνν̄)[C]
Empirical audit
closed CKM point→BR(K+ ⁣→π+ννˉ)\text{closed CKM point} \to \mathrm{BR}(K^+\!\to\pi^+\nu\bar\nu)
9.45×10⁻¹¹ (K⁺) · 3.33×10⁻¹¹ (K_L)(9.6⁺¹·⁹₋₁·₈)×10⁻¹¹+0.08σNA62 2016–2024 (La Thuile 2026)NA62 / KOTO-II (ongoing)
β_BH[C]
Empirical audit
βBH(r)=16 c34 QeQmr2=1256π4QeQmr2\beta_{\mathrm{BH}}(r) = 16\,c_3^4\,\tfrac{Q_e Q_m}{r^2} = \tfrac{1}{256\pi^4}\tfrac{Q_e Q_m}{r^2}
16c₃⁴ = 1/(256π⁴) (achromatic, 1/r², sign-flip)band-to-band EVPA +0.9° (intercept nulls open)data-limitedEHT M87 2017 polarimetry (2023)ngEHT + GRMHD library (this decade)
ν₃ = (3/2)ν_u[O]
Empirical audit
ladder step 32 from Nfam=3⇒ν3=32νu\text{ladder step } \tfrac{3}{2} \text{ from } N_{\mathrm{fam}} = 3 \Rightarrow \nu_3 = \tfrac{3}{2}\nu_u
tooth at 661.5/414/252/363 Hz, no integer linesno tooth, no integer line (~0σ, trials-corrected; archival scan 2026-07)well-powered null; 3σ limits 0.53–3.06% rms in all four sourcesRXTE PCA archive scan, 77 ObsIDs (Belloni+ 2012, RM06, Motta+ 2014–2022 epochs)eXTP-class sensitivity below the 0.5–3.1% rms limits
Reading rule. Every value is repo-documented — the measured central values and deviations come from the v307 data watchdog and the standalone experiments/tree (current bests: CODATA 2022, NuFIT 6.0, ACT DR6, Planck 2018, PDG 2024, BK18, DESI DR2, NA62 2026); the “decisive” column is the forward kill-test board (v321), whose timelines are the publicly stated experimental targets, not TFPT outputs. Deviations are colour-graded by the live empirical finding; no readout is upgraded by data proximity.
Committed kill criteria

Each output has a single sufficient kill condition

The solar angle θ₁₂ (JUNO), the tensor ratio r (CMB-S4), neutrino ordering, the strong-CP null, dark-energy w, the EM fixed point, the E₈ glue, the flavor invariants, and no second Higgs — each row is sufficient on its own to falsify the construction. m_p/m_e is listed for honesty: it is explicitly not claimed as a compiler power.

Solar angle θ₁₂

Doc 2 · JUNO (live)
Numerical kill

A JUNO central value clearly away from sin²θ₁₂ ≈ 0.307 at high significance kills the seam-misalignment mechanism. JUNO has been taking data since August 2025 — this is the sharpest live test.

sin⁡2θ12≠13−φ02≈0.3067\sin^2\theta_{12} \neq \tfrac{1}{3} - \tfrac{\varphi_0}{2} \approx 0.3067

Tau mass m_τ — conditional

Doc 4 · Belle II / BES III (near-term)
Conditional frontier test (not frozen)

Conditional near-term test (NOT a frozen prediction; v99). If the Koide source→pole flow runs exactly N_fam = 3 steps (one per family), the τ mass is forced to m_τ = 1776.9427 MeV. The PDG-2025 world average 1776.93 ± 0.09 MeV sits +0.14σ from it; Belle II's most precise single measurement (1777.09 ± 0.08 ± 0.11 MeV, pseudomass endpoint, PRD 108 032006) is +1.1σ above. Separating the n = 3 step from n = 4 (1776.967) needs σ(m_τ) ≈ 0.01 MeV — about 10× today's precision (Belle II full dataset / BES III threshold scan). A stable m_τ away from the step value kills the integer-step reading; the firewalled Koide flow itself stays [C].

mτ=?1776.9427 MeV(t=Nfam=3)m_\tau \stackrel{?}{=} 1776.9427\,\text{MeV}\quad (t = N_{\mathrm{fam}} = 3)

Universal recovery comb ω = 2.58 — exploratory

experiments/ · v425 single-flow
Conditional frontier test (not frozen)

Exploratory cross-domain signature (NOT a frozen prediction). v425 shows the four frontier transfers are ONE native seam recovery semigroup, so the same log-periodic comb at ω = 2π/ln((3/2)⁶) = 2.583 (amplitude ~2%) should appear in every wide-range recovery channel. Searched on real data (CHIME baseband + FAST FRB tails, magnetar outbursts, GRB afterglows) → clean null so far: ms FRB bursts are too short / scattering-dominated, and the cleaner wide-ln(t) channels (Vela day-cadence ν(t) timing, stacked BH late-time tails at O5+) are still data-limited. A confirmed ω = 2.58 comb in ≥ 2 independent wide-range channels would be striking; its absence only disfavours the recovery-channel reading — the firewall keeps the compiler core untouched.

ω=2π/ln⁡ ⁣((3/2)6)=2.583\omega = 2\pi/\ln\!\big((3/2)^6\big) = 2.583

Tensor ratio r

Doc 1 · CMB-S4
Numerical kill

Any robust r ≳ 0.01 is incompatible with the R² branch on which M_Pl and A_s rest. The scalaron predicts r = 12/N★² ≈ 0.004, already below the BK18 bound r < 0.036.

r≳0.01  ⇒  R2 branch killedr \gtrsim 0.01 \;\Rightarrow\; R^2\text{ branch killed}

Spectral running α_s

Doc 1 · CMB-S4 + DESI + Euclid (v494)
Numerical kill

The R² attractor fixes the running parameter-free: α_s = −2/N★² = −r/6 ∈ [−8.0, −5.6]×10⁻⁴. Plateau potentials give α_s < 0, so any robust POSITIVE running at ≥5σ kills the branch — the P-ACT-LB record leg already has a positive central value (+0.0062 ± 0.0052, −1.33σ from TFPT): the watch channel.

αs>0 robust (≥5σ)  ⇒  plateau branch killed\alpha_s > 0 \text{ robust } (\geq 5\sigma) \;\Rightarrow\; \text{plateau branch killed}

Neutrino ordering / m_ββ

Doc 2 · LEGEND, nEXO
Numerical kill

Inverted ordering, or a large effective Majorana mass m_ββ, kills the Majorana branch. TFPT prefers normal ordering with a small m_ββ.

inverted ordering, or mββ≳10−2 eV\text{inverted ordering, or } m_{\beta\beta} \gtrsim 10^{-2}\,\text{eV}

Leptonic CP phase δ_PMNS

Origin Theory · DUNE / Hyper-K (v320)
Structural kill

The two CP phases are Galois-locked: both are powers of one hexagonal unit ρ = ζ₆ of the family factor, so δ_PMNS = arg(ρ⁴) = δ_CKM,lead + π = 240° (the previously assigned value is now a forced relation to the measured quark phase). A measured δ_PMNS robustly away from 240° (beyond the sub-leading budget, >3σ at DUNE/Hyper-K/JUNO) falsifies the whole Galois-CP organisation.

δPMNS≠δCKMlead+π=240∘\delta_{\mathrm{PMNS}} \neq \delta_{\mathrm{CKM}}^{\mathrm{lead}} + \pi = 240^\circ

Strong CP θ_eff

Doc 2 · PSI nEDM
Identity / theorem kill

A solid neutron-EDM signal above the SM background falsifies the structural cancellation. θ_eff = 0 follows from γ₅-Hermiticity, polar structure and the sheet involution plus reflection positivity.

θeff≠0  ⇒  construction killed\theta_{\mathrm{eff}} \neq 0 \;\Rightarrow\; \text{construction killed}

Dark-energy w

Doc 1 · DESI
Cosmology kill

A robust w ≠ −1 kills the single-engine dark-energy readout, where Λ ∼ e⁻²ᵅ⁻¹ and H₀ ∼ √Λ come from the same exponential scale grammar.

w≠−1w \neq -1

EM fixed point α⁻¹

Doc 1 · CODATA
Numerical kill

F_U(1)(α) = 0 fails to admit a unique positive root, or the root drifts outside the declared interface uncertainty (currently ≈ 4 × 10⁻⁸ in α⁻¹, about 1.9σ of CODATA-2022).

FU(1)(α⋆)=0 has no/second root, or ∣Δα−1∣>ΔF_{U(1)}(\alpha_\star) = 0 \text{ has no/second root, or } |\Delta\alpha^{-1}| > \Delta

E₈ glue

Doc 1 (structural)
Identity / theorem kill

D₅ and A₃ fail to share the ℤ₄ discriminant, or the glue norms do not sum to the E₈ root norm 2. The whole compiler closure rests on this lattice fact.

disc⁡(D5)≠disc⁡(A3) or q(D5)+q(A3)≠2\operatorname{disc}(D_5) \neq \operatorname{disc}(A_3) \text{ or } q(D_5)+q(A_3) \neq 2

Flavor invariants

Doc 2 (structural)
Structural kill

A future global CKM/PMNS fit that cannot be carried by a residue matrix with det R = 8, principal 2-minors (2,3,5) and χ_R = t³ − 9t² + 10t − 8. Every load-bearing flavor number must live in an E₈ projection.

det⁡R≠8 or minors≠(2,3,5)\det R \neq 8 \text{ or } \mathrm{minors} \neq (2,3,5)

No second Higgs

Doc 1 (structural)
Structural kill

Robust discovery of a second light seam-even Higgs doublet. The carrier index forces N_Φ = g_car − |μ₄| = 1; an additional doublet kills it.

NΦ=1⇒no second light doubletN_\Phi = 1 \Rightarrow \text{no second light doublet}

m_p/m_e — not claimed

Doc 4 (honesty)
Structural kill

The proton/electron ratio is explicitly NOT claimed as a compiler power. It is a cross-sector QCD/EW ratio, deliberately not forced onto the ladder — it only fails if mis-asserted as a compiler power.

mp/me is [A], not a compiler powerm_p/m_e \text{ is } [\mathrm{A}],\ \text{not a compiler power}

Status discipline

Cross-cutting
Structural kill

A claim is promoted past the grade its document carries, an empirical input enters undeclared, or the text disagrees with the machine-checked ledger. The ledger always wins.

text≠status ledger\text{text} \neq \text{status ledger}
No-knobs audit

Inputs allowed, inputs forbidden, free knobs

A claim of 'no fitted constants' is only as strong as the audit table behind it. For each TFPT output the matrix records the inputs the construction may use (the two axioms and their consequences), the inputs it explicitly may not use, the number of free parameters available for absorption, the single condition that would falsify the row, and a link to run the check live. The free-knob count is the bar to clear.

No-knobs audit — what is allowed, what is forbiddenFree knobs: 0
OutputInputs allowedInputs forbiddenFree knobsKill conditionVerify
α⁻¹(0)c₃ = 1/(8π), b₁ = 41/10, the word-lengths Σ L + N_Φ = 41, the exact seam opening φ_seam(α)Fitting against CODATA / atom-recoil values; freezing φ_seam at φ₀ inside the root equation0No unique positive root of F_U(1)=0, or the root drifts outside the declared interface uncertainty (~1.9σ).Run it
sin²θ₁₂ (solar angle)The seed φ₀ and the glue norm q(A₃) = 3/4 (seam misalignment ε = (3/4)φ₀)NuFIT central value pre-loaded as input0A JUNO central value clearly away from 0.307 at high significance.Run it
sin²θ₁₃ (reactor angle)The seed φ₀ and the carrier trace e⁻⁵ᐟ⁶ (γ = 5/6)Oscillation-fit central value pre-loaded as input0A robust normal-ordering global-fit exclusion at the stated confidence level.Run it
δ_PMNS = 240° (leptonic CP phase)The hexagonal unit ρ = ζ₆ of the family Galois factor; δ_PMNS = arg(ρ⁴) = δ_CKM,lead + π, locked by ρ⁴ = −ρ (v320)A free leptonic CP phase fit to NuFIT0A measured δ_PMNS robustly away from δ_CKM,lead + π = 240° (>3σ at DUNE/Hyper-K).Run it
det R = 8, minors (2,3,5)The compiler residue matrix R = R(g_car, μ₄)Fitting R to a CKM/PMNS global fit0A CKM/PMNS global fit no residue matrix with det 8, minors (2,3,5) can carry.Run it
N_Φ = 1 (Higgs index)The carrier index, N_Φ = g_car − |μ₄| = 1Observed Higgs count used as primitive input0Robust discovery of a second light seam-even Higgs doublet.Run it
θ_eff = 0 (strong-CP null)γ₅-Hermiticity, polar structure, sheet involution + reflection positivityTuned θ-phase, hidden flavor-side cancellation0A solid neutron-EDM signal above the Standard-Model background.Run it
M_scal = c₃^(7/2) M̄ (scalaron)The seam power c₃⁷ = c₃^(Ω_adm − 10 b₁), exponent 7 = 48 − 41Fitting the scalaron mass to A_s0A scalaron mass incompatible with the seam power c₃⁷.Run it
n_s, r, A_s (inflation)The R² attractor + the seam-fixed scalaron mass + the e-fold count N★A free inflationary amplitude0 (amplitude) · N★ input (50–60)A robust r ≳ 0.01 kills the R² branch carrying M_Pl and A_s.Run it
β_rad = 0.2424° (birefringence)The determinant-line response, β_rad = φ₀/(4π)Calibration absorbed into the predicted angle0An externally calibrated β = 0 within tight error.Run it
Ω_b = 0.04894 (baryon density)β_rad via Ω_b = (4π − 1)β_radPlanck value used as primitive input0Robust inconsistency under the declared Planck comparison convention.Run it
m_a ≈ 23.8 µeV (axion DM)f_a = M_scal/128 and the closed misalignment θ_i = 170°Coupling rescaling to fit a haloscope window0 (decay-constant conjecture) · scenario-dependent (relic)Exclusion of the determinant-line axion window at the coupled sensitivity.Run it
Reading rule.For every row, the output value must be reproducible from the “allowed inputs” alone — without touching the “forbidden inputs”. The free-knob count is the number of parameters the theory may adjust to land on the listed value. Any row where this count exceeds zero is treated as a fit, not a prediction.