The Hibi--Li (2016) conjecture is fully resolved

Abstract

Hibi and Li conjectured that the order polytope 𝒪(P) of a finite poset P has at most as many faces of each dimension as its chain polytope 𝒞(P), and that equality in some dimension k with 1≤ k≤|P|−1 forces the two polytopes to be unimodularly equivalent. Fourier proved the inequality in a 2026 preprint. Let X be the five-element poset in which an element lies above two incomparable elements and below two other incomparable elements. Building on Fourier's proof, we show that if an n-element poset contains X as an induced subposet, then, for every k with 2≤ k≤ n−1, its chain polytope has at least binomn−3k−2 more k-dimensional faces than its order polytope, and that for every n≥5 some n-element poset containing X attains all these bounds at once. Since Hibi and Li showed that 𝒪(P) and 𝒞(P) are unimodularly equivalent exactly when P does not contain X, this proves the equality statement for these k. For k=1 the equality statement is false: 𝒪(X) and 𝒞(X) have equally many edges (Hibi, Li, Sahara and Shikama) but are not unimodularly equivalent. With Fourier's inequality, these results settle the conjecture. Every result is formally verified in Lean 4, with only published results and two results of Fourier's preprint as axioms.

The Hibi-Li Face-Number Conjecture
Ghislain Fourier
arXiv29 Oct 2026PreprintCC BY 4.0Full textCombinatorics & discrete mathematicsarXiv
The Fully Inhomogeneous p-Adic Littlewood Conjecture
Menny Aka, Alexander Gorodnik, Pankaj Vishe +1
arXiv3 Oct 2026PreprintarXiv licenceFull textDynamical systemsarXiv
Sign Propagation of Principal Minors in Fully Semimonotone Matrices and Stone's Conjecture
Sajal Ghosh
arXiv19 Oct 2026PreprintCC BY-NC-ND 4.0Full textAlgebra & number theory Numerical analysis & scientific computingarXiv
ARAZY'S CONJECTURE CONCERNING SCHUR MULTIPLIERS: REVISITED AND RESOLVED
Jinghao Huang, Fedor Sukochev
HAL7 Sep 2026PreprintPublisher termsFull textAnalysis & PDEsHALarXiv
Concavity-Free L^∞ Estimates and Stability for Fully Nonlinear Complex Equations
Ryan Mc Gowan
arXiv22 Oct 2026PreprintarXiv licenceFull textAnalysis & PDEs Geometry & topologyarXiv

Cite

BibTeX RIS

BibTeX

@misc{goel2026hibili,
  title = {{The Hibi--Li (2016) conjecture is fully resolved}},
  author = {Dave Goel},
  howpublished = {arXiv},
  year = {2026},
  month = oct,
  doi = {10.48550/arxiv.2610.40703},
  eprint = {2610.40703},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2610.40703},
}

RIS

TY  - UNPB
TI  - The Hibi--Li (2016) conjecture is fully resolved
AU  - Dave Goel
PY  - 2026
DA  - 2026/10/28
PB  - arXiv
DO  - 10.48550/arxiv.2610.40703
UR  - https://arxiv.org/abs/2610.40703
AB  - Hibi and Li conjectured that the order polytope 𝒪(P) of a finite poset P has at most as many faces of each dimension as its chain polytope 𝒞(P), and that equality in some dimension k with 1≤ k≤|P|−1 forces the two polytopes to be unimodularly equivalent. Fourier proved the inequality in a 2026 preprint. Let X be the five-element poset in which an element lies above two incomparable elements and below two other incomparable elements. Building on Fourier's proof, we show that if an n-element poset contains X as an induced subposet, then, for every k with 2≤ k≤ n−1, its chain polytope has at least binomn−3k−2 more k-dimensional faces than its order polytope, and that for every n≥5 some n-element poset containing X attains all these bounds at once. Since Hibi and Li showed that 𝒪(P) and 𝒞(P) are unimodularly equivalent exactly when P does not contain X, this proves the equality statement for these k. For k=1 the equality statement is false: 𝒪(X) and 𝒞(X) have equally many edges (Hibi, Li, Sahara and Shikama) but are not unimodularly equivalent. With Fourier's inequality, these results settle the conjecture. Every result is formally verified in Lean 4, with only published results and two results of Fourier's preprint as axioms.
ER  -

Abstract

Related

Cite