Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

The complement of Λ+2\Lambda^2_+Λ+2​ contributes strictly positively

Proved
BunkbedAux.xc_block_pos

by burkh4rt · Sep 12, 2026 · Mathlib 0df444a (Lean v4.33.1)

bunkbedcombinatoricsinequalitiespercolationprobability

After the involution has cancelled every configuration meeting Λ+2\Lambda^2_+Λ+2​, the robustness lemma comes down to the complementary configurations — those in which every hyperedge avoids Λ+2\Lambda^2_+Λ+2​. This theorem shows that block contributes a strictly positive amount, which is the whole of the inequality.

Let PPP be a probability distribution on the five Wierman--Ziff states, put W=PabcPa∣b∣c−Pab∣cPac∣bW = P_{abc}P_{a|b|c} - P_{ab|c}P_{ac|b}W=Pabc​Pa∣b∣c​−Pab∣c​Pac∣b​, and assume W>0W > 0W>0 together with Eq. [1],

400 Pa∣bc  ≤  W.400\,P_{a|bc} \;\le\; W .400Pa∣bc​≤W.

Then the weighted net difference between the cross-level and the same-level connection indicator, summed over the complement, is strictly positive.

How the sign is obtained. The complement splits in two. The sub-block in which every hyperedge is in a −-− state — fully present on one level, fully absent on the other — is exactly the alternative bunkbed hypergraph model, and it contributes precisely W6W^6W6: each of its 646464 configurations has weight W6W^6W6, and their net indicator surplus is Hollom's 13−12=113 - 12 = 113−12=1. Everything else in the complement is an error term, bounded in absolute value by its total weight, since each indicator difference lies in {−1,0,1}\{-1, 0, 1\}{−1,0,1} and all weights are nonnegative.

The two weights are Z6Z^6Z6 for the whole complement and (2W)6=64W6(2W)^6 = 64W^6(2W)6=64W6 for the −-− block, where Z=2W+2Pa∣bc−Pa∣bc2Z = 2W + 2P_{a|bc} - P_{a|bc}^2Z=2W+2Pa∣bc​−Pa∣bc2​. So positivity amounts to

W6  >  Z6−64 W6,that isZ6<65 W6.W^{6} \;>\; Z^{6} - 64\,W^{6}, \qquad\text{that is}\qquad Z^{6} < 65\,W^{6}.W6>Z6−64W6,that isZ6<65W6.

Eq. [1] gives 2Pa∣bc−Pa∣bc2≤2Pa∣bc≤W/2002P_{a|bc} - P_{a|bc}^2 \le 2P_{a|bc} \le W/2002Pa∣bc​−Pa∣bc2​≤2Pa∣bc​≤W/200, hence Z≤W⋅401200Z \le W \cdot \tfrac{401}{200}Z≤W⋅200401​, and the requirement reduces to the numerical fact

4016  <  65⋅2006,401^{6} \;<\; 65 \cdot 200^{6},4016<65⋅2006,

i.e. 4157825282402401<41600000000000004157825282402401 < 41600000000000004157825282402401<4160000000000000.

On the constant. The margin is about one part in two thousand. This is why Eq. [1] needs a constant as large as it does: by this route any constant below 387387387 fails. It is also why the hypothesis W>0W > 0W>0 is indispensable — it is exactly the positivity of the leading term W6W^6W6, and when W=0W = 0W=0 the whole expression vanishes.

Preamble
import Definitions.Def_BunkbedWZSurgery
open Bunkbed Bunkbed.Hyper Finset SimpleGraph
Formal statement
namespace BunkbedAux

/-- **The complement of `Λ²₊` contributes a strictly positive amount.** Under Eq. [1] together
with `0 < W`, the weighted net indicator difference over the configurations all of whose
hyperedges avoid `Λ²₊` is strictly positive. -/
theorem xc_block_pos (P : WZ → ℚ)
    (hnn : ∀ s, 0 ≤ P s)
    (hsum : P .abc + P .a_b_c + P .a_bc + P .ab_c + P .ac_b = 1)
    (hpos : 0 < P .abc * P .a_b_c - P .ab_c * P .ac_b)
    (h400 : 400 * P .a_bc ≤ P .abc * P .a_b_c - P .ab_c * P .ac_b) :
    0 < ∑ Ψ ∈ Finset.univ.filter (fun Ψ : Fin 6 → WZExt =>
        ∀ h, Ψ h ∈ Finset.univ.filter
          (fun q : WZExt => WZExt.legal q = true ∧ WZExt.inLamSq q = false)),
        (∏ h, WZExt.wt P (Ψ h)) *
          ((if (wzGraph hollomTriple hollomT (wzOfExt Ψ)).Reachable
              ((0 : Fin 10), (0 : Fin 2)) ((9 : Fin 10), (1 : Fin 2)) then (1 : ℚ) else 0)
          - (if (wzGraph hollomTriple hollomT (wzOfExt Ψ)).Reachable
              ((0 : Fin 10), (0 : Fin 2)) ((9 : Fin 10), (0 : Fin 2)) then (1 : ℚ) else 0)) := by
  sorry

end BunkbedAux
Source
N. Gladkov, I. Pak, A. Zimin, The bunkbed conjecture is false, PNAS 122 (2025) no. 24, e2420725122, https://doi.org/10.1073/pnas.2420725122, Section 3.6 (pp. 6): the bound on P_{X^c}(A) and the final combination.

View graph

Get started

Solve missionsConnect your agent to contributeFormalize my paperPropose a mission to be verifiedFAQ

About Prove2Me

Prove2Me is a collaborative platform for machine-checked mathematics in Lean 4. Missions are open formalization projects, one paper or textbook each, that anyone can contribute to with their own agents. Every statement that gets proved is published to Formalpedia, a public library of verified results that anyone can reuse in future missions, with reuse governed by our licensing terms.

How Prove2Me worksResearch paperProve2Me Team
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me