Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Counterexample.

Proved
TropicalLA.no_spread_lower_bound

by raver1975 · Sep 17, 2026 · Mathlib c5ea003 (Lean v4.30.0)

aether-catalogalgebra

Counterexample. For A = [[0,−3],[0,−3]] the constant vector is a tropical eigenvector, so its spread is 0 and λ(A) = 0; nevertheless every entry in the second column of every tropical power is at most −3. Hence the two-sided box bound with constant spread v is false: only the upper half of sub-conjecture 1 holds, and the lower bound genuinely needs a constant depending on the entries of A.

theorem TropicalLA.no_spread_lower_bound:
    ∃ (A : Matrix (Fin 2) (Fin 2) ℝ) (v : Fin 2 → ℝ),
      IsTropEigen A (maxCycleMean A) v ∧ spread v = 0 ∧
      ∀ (m : ℕ) (i : Fin 2), tpow A m i 1 ≤ ((m : ℝ) + 1) * maxCycleMean A - 3 := by sorry

Formalization Note Transplanted verbatim from the Aether Catalog source Algebra/TropicalLinearAlgebra/TropicalCyclicity.lean; the statement is byte-identical to the source declaration, elaborated with autoImplicit disabled in the platform environment.

Preamble
-- Thm stub generated from Algebra/TropicalLinearAlgebra/TropicalCyclicity.lean
import Mathlib
import Definitions.Def_Algebra_TropicalLinearAlgebra_TropicalCyclicity
import Definitions.Def_Algebra_TropicalLinearAlgebra_TropicalEigenvalue
import Definitions.Def_Algebra_TropicalLinearAlgebra_TropicalGelfand
import Definitions.Def_Algebra_TropicalLinearAlgebra_TropicalMatrix
import Definitions.Def_Algebra_TropicalLinearAlgebra_TropicalPerronFrobenius
/-
# Uniform boundedness of normalised tropical powers

`FUTURE_DIRECTIONS.md` (sub-conjecture 1, extracted from conjecture C1 on cyclicity)
proposed that *every* entry of `A^{⊗(m+1)} − (m+1)·λ` lies in the box `[−C, C]` with
`C = spread(v)` the spread of a tropical eigenvector.  This file settles that
sub-conjecture, in the sharp form:

* `tpow_le_spread` — the **upper** half of the conjecture is true exactly as stated:
  `tpow A m i j ≤ (m+1)·λ + spread v` for every eigenvector `v`;
* `no_spread_lower_bound` — the **lower** half is *false*: for
  `A = [[0,−3],[0,−3]]` the constant vector is an eigenvector (spread `0`) while the
  whole second column of every tropical power stays `3` below `(m+1)·λ`;
* `exists_uniform_entry_bound` — but the *boundedness* statement survives with a larger,
  explicitly computable constant: with `q ≤ n` the length of a critical cycle and
  `a = min_{i,j} A i j`,
  `|tpow A m i j − (m+1)·λ| ≤ spread v + (1+q)·|a − λ|` for all `m ≥ q + 1`.
  The lower bound is obtained by *constructing* long walks: one step into a critical
  node, `k` turns around the critical cycle, and a short tail of length `r ≤ q`.
* `tendsto_tpow_div` — consequently the entrywise Gelfand formula holds:
  `tpow A m i j / (m+1) → λ` for *every* pair `(i,j)`, a strengthening of the
  largest-entry version `tendsto_specNorm_div`.

The technical engine is the semigroup law `tpow_add : A^{⊗(a+b+2)} = A^{⊗(a+1)} ⊗ A^{⊗(b+1)}`
together with the critical cycle produced by tropical Perron–Frobenius.
-/

open TropicalLA

open Filter Topology

variable {ι : Type*} [Fintype ι] [Nonempty ι]

/-! ## The semigroup law for tropical powers -/



/-! ## Smallest entry -/




/-! ## Upper bound from an eigenvector -/




/-! ## Powers of a critical cycle -/


/-! ## The uniform two-sided entrywise bound -/




/-! ## The spread is *not* a lower bound: refutation of the naive box conjecture -/
Formal statement
theorem TropicalLA.no_spread_lower_bound:
    ∃ (A : Matrix (Fin 2) (Fin 2) ℝ) (v : Fin 2 → ℝ),
      IsTropEigen A (maxCycleMean A) v ∧ spread v = 0 ∧
      ∀ (m : ℕ) (i : Fin 2), tpow A m i 1 ≤ ((m : ℝ) + 1) * maxCycleMean A - 3 := by sorry
Source
https://github.com/paulklemstine/Lean/blob/53c2925a02/Catalog/Algebra/TropicalLinearAlgebra/TropicalCyclicity.lean#L262

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