Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Theorem 6.7 (Motzkin–Straus) — 1 − 1/ω(G) = 2 max over the simplex of Σ_{ij∈E} x_i x_j

Proved
TensorNP.SpectralNorm.theorem_6_7

by mikedeng1 · Oct 7, 2026 · Mathlib 0df444a (Lean v4.33.1)

clique-numbergraph-theorymotzkin-strausp2o-batch-pfp2bp2o-gran-per-chapterp2o-plan-paperp2o-v1quadratic-optimization

Let G=(V,E)G=(V,E)G=(V,E) be a simple graph on v≥1v\ge1v≥1 vertices with clique number ω(G)\omega(G)ω(G) (the number of vertices of a largest clique), and let Δv={x∈R≥0v:∑i=1vxi=1}\Delta_v=\{\mathbf x\in\mathbb R^v_{\ge0}:\sum_{i=1}^v x_i=1\}Δv​={x∈R≥0v​:∑i=1v​xi​=1} be the standard simplex. Then

1−1ω(G)=2⋅max⁡x∈Δv∑{i,j}∈Exixj,1-\frac1{\omega(G)} = 2\cdot\max_{\mathbf x\in\Delta_v}\sum_{\{i,j\}\in E}x_ix_j ,1−ω(G)1​=2⋅x∈Δv​max​{i,j}∈E∑​xi​xj​,

where the sum runs over the unordered edges of GGG, and the maximum is attained.

This classical theorem of Motzkin and Straus (1965) links the combinatorial quantity ω(G)\omega(G)ω(G) to a quadratic optimization problem over the simplex; it is the source of the value of MlM_lMl​ in display (22).

Formalization Note The vertex set is Fin v and the clique number is Mathlib's SimpleGraph.cliqueNum. The statement is IsGreatest of the image of Δv\Delta_vΔv​ (Mathlib's stdSimplex ℝ (Fin v)) under x↦2∑{i,j}∈Exixj\mathbf x\mapsto 2\sum_{\{i,j\}\in E}x_ix_jx↦2∑{i,j}∈E​xi​xj​, which asserts both that the value 1−1/ω(G)1-1/\omega(G)1−1/ω(G) is attained and that it is an upper bound. The hypothesis v≥1v\ge1v≥1 (so ω(G)≥1\omega(G)\ge1ω(G)≥1) excludes the empty graph, for which 1/ω1/\omega1/ω is not defined.

Preamble
import Mathlib
Formal statement
namespace TensorNP.SpectralNorm

/-- **Theorem 6.7 (Motzkin–Straus).** Let `G` be a simple graph on `v ≥ 1` vertices with clique
number `ω(G)`. Over the standard simplex `Δ_v`, the maximum of `2 Σ_{{i,j} ∈ E} x_i x_j` (sum over
the unordered edges) is `1 − 1/ω(G)`. -/
theorem theorem_6_7 {v : ℕ} (hv : 0 < v) (G : SimpleGraph (Fin v)) [DecidableRel G.Adj] :
    IsGreatest
      ((fun x : Fin v → ℝ =>
          2 * ∑ s ∈ G.edgeFinset, Sym2.lift ⟨fun i j => x i * x j, fun i j => mul_comm _ _⟩ s) ''
        stdSimplex ℝ (Fin v))
      (1 - 1 / (G.cliqueNum : ℝ)) := by sorry

end TensorNP.SpectralNorm
Source
Hillar & Lim, Most Tensor Problems Are NP-Hard, J. ACM 60(6) (2013), Art. 45 (final version, arXiv:0911.1393v5), p. 0:22, Theorem 6.7 (Motzkin–Straus 1965)
Human review
  • Endorsed by Shuze Chen · Oct 8, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 8, 2026

    Confirmed by the mission captain (proposal self-audit).

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