AFTD Auto-Formalizing Theoretical Domains

five_point_zero_sum_unit_atom

theoremverified

The all-one multiplicity vector is an atom of the integer zero-sum kernel of (1,1),(1,0),(0,1),(-2,0),(0,-2): any nonnegative integer zero-sum subvector with all coordinates at most one is zero or the all-one vector.

Statement

theorem five_point_zero_sum_unit_atom :
    ∀ y : Fin 5 → ℤ, (∀ i, 0 ≤ y i) → (∀ i, y i ≤ 1) →
      y 0 + y 1 = 2*y 3 → y 0 + y 2 = 2*y 4 →
      y = 0 ∨ y = ![1,1,1,1,1]
Source
Supporting lemma or explicit route obstruction for OP-197 / arXiv:2610.16562 Conjecture1, deep session2. The abstract two-ray fact and finite obstruction are not claimed novel or research-level; no part is the full box conjecture. The five-point kernel has columns (1,1),(1,0),(0,1),(-2,0),(0,-2).
Verified
28 Oct 2026
Axioms
Classical.choiceQuot.soundpropext
Used by 1

The proof, explained

The proof proceeds by case analysis on the first coordinate $y_0 \in \{0, 1\}$. If $y_0 = 0$, the relations $y_1 = 2y_3$ and $y_2 = 2y_4$ together with $0 \le y_i \le 1$ force all coordinates to be zero. If $y_0 \ne 0$, then $y_0 = 1$, and parity in $1 + y_1 = 2y_3$ and $1 + y_2 = 2y_4$ forces all remaining coordinates to be 1. In both branches, checking every coordinate reduces to linear integer arithmetic, handled automatically by `omega`.

  1. Assume $y \in \mathbb{Z}^5$ satisfies $0 \le y_i \le 1$ for all $i$ along with the two zero-sum conditions, and instantiate the bounds for each coordinate.

    We introduce the vector $y$, the bounds $0 \le y_i \le 1$, and the linear relations $y_0 + y_1 = 2y_3$ and $y_0 + y_2 = 2y_4$. We then instantiate the bounds for each specific index $i \in \{0, 1, 2, 3, 4\}$, adding the explicit inequalities $0 \le y_i$ and $y_i \le 1$ to the local context.

    lines 9–11

      intro y hy hyb hz1 hz2
      have h0 := hy 0; have h1 := hy 1; have h2 := hy 2; have h3 := hy 3; have h4 := hy 4
      have hb0 := hyb 0; have hb1 := hyb 1; have hb2 := hyb 2; have hb3 := hyb 3; have hb4 := hyb 4
  2. The argument splits into two cases: $y_0 = 0$ or $y_0 \ne 0$.

    We perform a case analysis on whether the first coordinate $y_0$ is zero.

    line 12

      by_cases hh : y 0 = 0
  3. If $y_0 = 0$, then $y$ is the zero vector.

    Selecting the left disjunct ($y = 0$), it suffices by function extensionality to verify $y_i = 0$ for each coordinate. We exhaust all five coordinates $i \in \{0, 1, 2, 3, 4\}$ with `fin_cases`; in each case, Presburger arithmetic (`omega`) deduces $y_i = 0$ because $y_0 = 0$ together with $y_1 = 2y_3$, $y_2 = 2y_4$, and $y_i \in \{0, 1\}$ forces every coordinate to vanish.

    line 13

      · left; funext i; fin_cases i <;> dsimp at * <;> omega
  4. If $y_0 \ne 0$, then $y$ is the all-one vector.

    Selecting the right disjunct ($y = ![1, 1, 1, 1, 1]$), we show by function extensionality that $y_i = 1$ for each coordinate. Case analysis on $i \in \{0, 1, 2, 3, 4\}$ leaves five scalar goals, all closed by Presburger arithmetic (`omega`): since $0 \le y_0 \le 1$ and $y_0 \ne 0$, we have $y_0 = 1$, which forces $1 + y_1 = 2y_3$ and $1 + y_2 = 2y_4$, whose only binary solutions are $y_1 = y_2 = y_3 = y_4 = 1$.

    line 14

      · right; funext i; fin_cases i <;> dsimp at * <;> omega

Written by a model (gemini-3.8-flash) from the Lean proof and the goal states Lean reported at each step. Lean verified the proof; this explanation is a reading aid and is not itself machine-checked.

Lean source view module on GitHub

open Finsetset_option maxHeartbeats 1000000set_option maxRecDepth 4000theorem five_point_zero_sum_unit_atom :    ∀ y : Fin 5 → ℤ, (∀ i, 0 ≤ y i) → (∀ i, y i ≤ 1) →      y 0 + y 1 = 2*y 3 → y 0 + y 2 = 2*y 4 →      y = 0 ∨ y = ![1,1,1,1,1] := by  intro y hy hyb hz1 hz2  have h0 := hy 0; have h1 := hy 1; have h2 := hy 2; have h3 := hy 3; have h4 := hy 4  have hb0 := hyb 0; have hb1 := hyb 1; have hb2 := hyb 2; have hb3 := hyb 3; have hb4 := hyb 4  by_cases hh : y 0 = 0  · left; funext i; fin_cases i <;> dsimp at * <;> omega  · right; funext i; fin_cases i <;> dsimp at * <;> omega
See it in the graphDiscuss this resultChallenge it