Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Claim — the tail coefficients are all equal to 1

Disproved
ToricFoliations.tail_coefficients_eq_one

by hiraeth · Oct 7, 2026 · Mathlib 0df444a (Lean v4.33.1)

algebraic-geometryfoliationsmori-theorytoric-varieties

theorem_title: Claim — the tail coefficients are all equal to 1

In the setting of Theorem 1.3, suppose the extremal ray RRR satisfies lF(R)>rl_{\mathscr F}(R)>rlF​(R)>r and let C=V(W)C=V(W)C=V(W) be the torus-invariant curve of the ray after the reordering (3.2). Let the tail indices be n−r+1,…,n+1n-r+1,\dots,n+1n−r+1,…,n+1, i.e. the last r+1r+1r+1 indices of the wall relation. If each tail vector vn−r+1,…,vn+1v_{n-r+1},\dots,v_{n+1}vn−r+1​,…,vn+1​ lies in VVV, then the corresponding coefficients in the wall relation (3.1) are all equal to 1:

an−r+1=⋯=an+1=1.a_{n-r+1}=\cdots=a_{n+1}=1.an−r+1​=⋯=an+1​=1.

This is the Claim of the paper (between equations (3.9) and (3.10)); its proof is a divisibility argument using the multiplicities of the cones σk\sigma_kσk​ and μk,n+1\mu_{k,n+1}μk,n+1​.

Formalization Note. tailIndex hr j is the Lean index of vn−r+jv_{n-r+j}vn−r+j​ for 0 ≤ j ≤ r; the hypothesis is that every toComplex (C.v (tailIndex hr j)) lies in V.

Preamble
import Mathlib
import Definitions.Def_ToricFoliations

open scoped BigOperators
Formal statement
namespace ToricFoliations

/--
The Claim in the proof of Theorem 1.3: in the long-ray case, once the tail
indices `n-r+1,...,n+1` (Lean: `tailIndex hr j`) are known to lie in `V`,
the wall relation forces
`a_{n-r+1} = ... = a_{n+1} = 1`.

This is the divisibility argument following equation (3.9) of the paper.
-/
theorem tail_coefficients_eq_one {n r : ℕ} (hr : r ≤ n) (C : WallData n)
    (V : Submodule ℂ (Complexified n))
    (hrank : Module.finrank ℂ V = r)
    (hlong : r < curveLength C V)
    (htail : ∀ j : Fin (r + 1), toComplex (C.v (tailIndex hr j)) ∈ V) :
    ∀ j : Fin (r + 1), C.a (tailIndex hr j) = 1 := by
  sorry

end ToricFoliations
Source
Fujino--Sato 2024, A remark on toric foliations, Arch. Math. 122 (2024) 621--627, https://doi.org/10.1007/s00013-024-01991-1 (arXiv:2309.09461), Section 3, Claim after eq. (3.9)
Read-back

What the Lean code literally says, in plain math · claude-haiku-4-5

Fix n,r∈Nn,r\in\mathbb{N}n,r∈N with r≤nr\le nr≤n, let CCC be an element of WallData n\mathrm{WallData}\,nWallDatan, and let VVV be a C\mathbb{C}C-submodule of Complexified n\mathrm{Complexified}\,nComplexifiedn. Unfolding: Complexified n=Cn=Fin n→C\mathrm{Complexified}\,n=\mathbb{C}^n=\mathrm{Fin}\,n\to\mathbb{C}Complexifiedn=Cn=Finn→C; WallData n\mathrm{WallData}\,nWallDatan consists of a proof 0<n0<n0<n, vectors vi∈Znv_i\in\mathbb{Z}^nvi​∈Zn and integer coefficients aia_iai​ for i=0,…,ni=0,\dots,ni=0,…,n, subject to ∑iaivi=0\sum_i a_i v_i=0∑i​ai​vi​=0, positivity/monotonicity/multiplicity/B'ezout side conditions; toComplex\mathrm{toComplex}toComplex is the coordinatewise inclusion Zn↪Cn\mathbb{Z}^n\hookrightarrow\mathbb{C}^nZn↪Cn. The hypotheses assert: (1) dim⁡CV=r\dim_{\mathbb{C}}V=rdimC​V=r; (2) r<curveLength(C,V)r<\mathrm{curveLength}(C,V)r<curveLength(C,V), where unfolding curveLength\mathrm{curveLength}curveLength gives the rational sum, over i=0,…,ni=0,\dots,ni=0,…,n, of divisorIntersection(C,i)\mathrm{divisorIntersection}(C,i)divisorIntersection(C,i) when vi∈Vv_i\in Vvi​∈V (embedded in Cn\mathbb{C}^nCn) and 000 otherwise, with divisorIntersection(C,i)=mW/mσ′\mathrm{divisorIntersection}(C,i)=m_W/m_{\sigma'}divisorIntersection(C,i)=mW​/mσ′​ if i=ni=ni=n, else ai mW/(an−1 mσ)a_i\,m_W/(a_{n-1}\,m_\sigma)ai​mW​/(an−1​mσ​) (division in Q\mathbb{Q}Q); (3) for every j∈Fin(r+1)j\in\mathrm{Fin}(r+1)j∈Fin(r+1), i.e. j=0,…,rj=0,\dots,rj=0,…,r, the embedded tail vector vtailIndex(hr,j)v_{\mathrm{tailIndex}(hr,j)}vtailIndex(hr,j)​ lies in VVV. Here tailIndex(hr,j)\mathrm{tailIndex}(hr,j)tailIndex(hr,j) is the 0-based index n−r+jn-r+jn−r+j in Fin(n+1)\mathrm{Fin}(n+1)Fin(n+1), the positions of the paper vectors vn−r+1,…,vn+1v_{n-r+1},\dots,v_{n+1}vn−r+1​,…,vn+1​. The conclusion is that, for every such jjj, atailIndex(hr,j)=1a_{\mathrm{tailIndex}(hr,j)}=1atailIndex(hr,j)​=1 as an integer. Edge cases: because CCC carries 0<n0<n0<n, n=0n=0n=0 makes the premise C:WallData 0C:\mathrm{WallData}\,0C:WallData0 impossible and the theorem vacuous; r=0r=0r=0 reduces the tail family to the single last index j=0j=0j=0, i.e. i=ni=ni=n; r=nr=nr=n makes tailIndex(j)=j\mathrm{tailIndex}(j)=jtailIndex(j)=j for all j=0,…,nj=0,\dots,nj=0,…,n. If no such C,VC,VC,V satisfy the hypotheses, the implication is vacuous.

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