Different four-label laws have identical first-order block data
ProvedFormalCapacity.Finite.n4_sameFirstOrderfinitecombinatoricsformalcapacitylinearalgebrapartition-incidence
On four labels, let be uniform on and , and let be uniform on and . For every subset of the labels, with rational exact-block indicators,
Thus these explicit laws cannot be distinguished by first-order block coordinates. The assertion is equality of block statistics, not equality of the partition laws or the existence of a refinement coupling. The quantifier includes the empty subset, for which both coordinates vanish.
Preamble
import Mathlib import Definitions.Def_capacityFourLabelAtoms set_option autoImplicit false /-! # The four-label correlation obstruction This file kernel-checks the signed square `12|34 + 1|2|3|4 = 1|2|34 + 12|3|4` at every block-incidence coordinate. The four atoms are represented by their finite sets of blocks. `n4Atom_isPartition` verifies that each label belongs to exactly one nonempty block, so this concrete representation does not hide a partition-validity assumption. -/ open FormalCapacity.Finite open Finset
Formal statement
theorem FormalCapacity.Finite.n4_sameFirstOrder (B : Finset Label4) :
n4TwoAtomAverage .plusCoarse .plusDiscrete
(fun P ↦ P.blockIndicator B) =
n4TwoAtomAverage .minusLeft .minusRight
(fun P ↦ P.blockIndicator B) := by
sorry
Source
The four-label correlation threshold, Proposition 1.1 and (1.3). Unpublished research note (2026), N4_BLOCK_CORRELATION_NOTE.md, SHA-256 a9b67c132ca0e0af807fb917ca2322833b84e162c45a4da7329ee88fbe4f1c83. Exact formal source: formal_capacity/FormalCapacity/Finite/N4.lean, lines 90–97, declaration FormalCapacity.Finite.n4_sameFirstOrder, source-file SHA-256 280e7a5db975df324036a6fa34b42c689519d3d508ba10e85a18645f77ddeb26. Local source archive; no public repository URL or commit is asserted. Upload checked with Lean 4.33.1 and Mathlib 0df444a360eaa60ab8c11dca51a86af692955474.