Two-neighborly simplicial 0/1-polytopes
Abstract
A 0/1-polytope is the convex hull of a subset of the vertices of the unit cube. Ziegler proved that a simplicial 0/1-polytope of dimension d has at most 2d vertices and asked whether the equality case is always a cross polytope; Kaibel and Pokutta recently answered this in the negative with an explicit example in dimension 7, and asked in turn whether a simplicial 0/1-polytope can be 2-neighborly at all. We answer their question affirmatively by exhibiting simplicial 2-neighborly 0/1-polytopes in dimensions 4,5,6,7, with 6,8,10,12 vertices respectively; the three of dimension at least 5 have, in addition, no cube-antipodal pair of vertices, which is the stronger form in which the question is posed. Each example is certified twice and independently: through a complete facet enumeration, and through an explicit integer separating functional for every pair of vertices. We complement this with complete censuses of every subset of {0,1}³ and of {0,1}⁴, which show that in those dimensions the largest 2-neighborly simplicial example has 2d-2 vertices and that none of the 2d-vertex examples is 2-neighborly, and we recompute the invariants of Kaibel and Pokutta's dimension-7 examples from their definitions rather than from their tables. Every theorem and proposition below is machine-checked in Lean 4.
Open review
This founding-collection manuscript received AI review before publication. Independent human review is open. Submitted reviews enter editorial screening; submitting a review does not change this paper’s status. Contribute an assessment of specific claims, a reproduction, or a correction for editorial screening.
Archived files
- Version 1 · current (opens in a new tab)
Source snapshot 2026-08-30 15:34 UTC
File fingerprint
a35d9b3a26cb70f737e2959180444413304de734d6ae79e8920e4c6355712b95
Claim ledger
Stated results
ZP1known data2026-08-22
Kaibel-Pokutta's counterexample replayed: a simplicial, not centrally symmetric 0/1 7-polytope with 14 vertices and 136 facets, f- and h-vectors recomputed
ZP2known data2026-08-22
The other four B₇-orbits and the coordinate swap: Table 3's two combinatorial types and Remark 3.4's cross polytope
ZP3routine2026-08-22
Complete censuses of every subset of the 3-cube and the 4-cube: Ziegler's question is true at d = 3 and d = 4, the bound f₀ <= 2d is sharp, and nothing at |V| = 2d is 2-neighborly
ZP4candidate2026-08-23
2-neighborly simplicial 0/1-polytopes exist: witnesses at d = 4, 5, 6, 7, each certified twice
ZP5routine2026-08-22
The geometry the integer certificates stand on, kernel-clean over the reals
ZP6routine2026-08-22
Negative controls: the cube is not simplicial, the cross polytope is not 2-neighborly, and the three conditions are separated
Provenance
- Generated by
- Machina Mathematica
- Released by
- Korea Superintelligence Labs
- Source context
- Source: arXiv:2606.31640, Volker Kaibel and Sebastian Pokutta, *A Counterexample to Ziegler's Cross-Polytope Conjecture for Simplicial 0/1-Polytopes*, v1 30 June 2026, v2 2 July 2026.
- Snapshot
- 2026-09-07 03:53 UTC
- Ledger commit
801848d7