Skip to content

CLAIMS.md

Latest commit

 

History

History
68 lines (53 loc) · 4.15 KB

CLAIMS.md

File metadata and controls

68 lines (53 loc) · 4.15 KB
Raw

Claims and their status

TFPT's discipline is that nothing is marked closed that is not machine-verified, and no dimensionful quantity is claimed as a derivation from pure numbers. The authoritative, per-claim status is the versioned ledger verification/status_ledger.csv; this page is the human-readable summary. When in doubt, the ledger wins.

The headline claims and their epistemic status

Not everything here has the same epistemic status — and saying so is the point. The five strongest statements, made explicit:

Claim Result Status
E8 closure (D5 ⊕ A3) + μ4 ≅ E8, 240 roots, glue index 4 Exact — machine-proven lattice identity [E]
Number of generations N_fam = 3 Exact within the compiler — = rank A3 = dim H₁(P¹∖μ4) [E]
Fine-structure constant α⁻¹ = 137.0359992 Exact numerical identity — unique root of the boundary Ward identity, interval-verified [E] (1.9σ from CODATA-2022)
Physical origin of the seam the 1/(8π) area law / SEAM.EQUIV.01 Open — closed modulo a cited theorem, not proven end-to-end [O]
Cosmic birefringence β = φ₀/(4π) ≈ 0.2424° Falsifiable prediction — frozen, decided by CMB polarimetry [X]

Read this as: the discrete/algebraic compiler is closed and exact; the physical identification of the boundary object is the honest open frontier; and there are sharp numbers that observation can kill. See FALSIFICATION.md for the kill tests and OPEN_PROBLEMS.md for the frontier.

Status markers

Used consistently across all documents and the ledger. The documents show a simplified four-class display marker; the ledger keeps the fine-grained per-claim type (Axiom / Formal / Lattice / Numerical / Identity / Physical), so no fidelity is lost.

Display marker Meaning Ledger fine types it covers
[E] exact / machine-proven Identity, Lattice (Lie/lattice), Formal (Lean), Numerical
[C] conditional (holds under named hypotheses) Physical, bridge, readout
[O] open / axiom (declared input or genuine gap) Axiom
[X] falsifiable kill test —

The ledger is append-only and versioned: superseded rows are marked active=false with a canonical_status pointer, so the current authoritative status of any claim is unambiguous. A status marker is never silently upgraded.

The four layers (where each kind of claim lives)

Layer Content Status
1. Closed compiler E8 glue, carrier, α⁻¹, (R,K,Q,L), lepton/quark ratios [I]/[L]/[N]
2. Protected IR physics R+R², admissible gapped transfer sector; the boundary QFT as one relative object [I]/[P]
3. Anchors π, one dimensionful induced-gravity scale, U_point absolute amplitude norm [A] (declared, not "missing")
4. Interfaces m_p/m_e, η_B (leptogenesis), Koide, axion relic, full ambient QG measure [P]/[A]

How every claim is kept honest

Every script cited in verification/run_all.py is also cited inline in the theory documents via \veri{vN_*.py} (enforced in both directions by verification/audit_sync.py), and the status heatmap is generated directly from status_ledger.csv. So the papers, the suite, the ledger and the website stay in lock-step: a claim cannot appear in a paper without a machine check behind it, and a check cannot exist without a ledger row typing its status.

  • 472 numbered claim checks (v1_*.py … v472_*.py), one per claim cluster.
  • Independent Wolfram path (116/116 base, 368/368 extension) re-derives the exact algebraic/identity/lattice results on a second engine.
  • Lean 4 machine-formalises the carrier-rigidity core ([F]): no sorry/admit; every headline theorem's #print axioms returns only the three standard kernel axioms.
  • Red Team (redteam/run_redteam.py) actively tries to break the five load-bearing reductions.

See VERIFICATION.md to reproduce any of this.