Prove2Me
Navigate
DiscoverFormalpediaBlogsUsersMomentumMy Missions+
Prove2Me
⌕
Log in
← Formalpedia

Zero net power   ⟺  \iff⟺ every link coupling is symmetric, on a qqq-dimensional lattice (L≥3L \ge 3L≥3)

Proved
PassivityTorus.passive_iff_symm

by ShapeZero · Sep 24, 2026 · Mathlib 0df444a (Lean v4.33.1)

latticelinear-algebramatrices

Let qqq and ddd be natural numbers and L≥3L \ge 3L≥3. On the periodic cubic lattice (Z/LZ)q(\mathbb{Z}/L\mathbb{Z})^q(Z/LZ)q, let each link (x,a)(x,a)(x,a) carry a real d×dd\times dd×d matrix W(x,a)W(x,a)W(x,a), with total power

PW(v)=∑x∑a=0q−1v(x)⋅(W(x,a) v(x+ea)−W(x−ea,a) v(x−ea)).P_W(v) = \sum_{x} \sum_{a=0}^{q-1} v(x) \cdot \Bigl( W(x,a)\, v(x+e_a) - W(x-e_a,a)\, v(x-e_a) \Bigr).PW​(v)=x∑​a=0∑q−1​v(x)⋅(W(x,a)v(x+ea​)−W(x−ea​,a)v(x−ea​)).

Then

(PW(v)=0 for every v)  ⟺  (W(x,a)T=W(x,a) for every x,a).\bigl( P_W(v) = 0 \ \text{for every } v \bigr) \iff \bigl( W(x,a)^{\mathsf T} = W(x,a) \ \text{for every } x, a \bigr).(PW​(v)=0 for every v)⟺(W(x,a)T=W(x,a) for every x,a).

A per-link neighbour coupling on a lattice of any dimension does no net work for every motion exactly when every link matrix is symmetric. It does not assert that the coupling commutes with any complex structure.

Preamble
import Mathlib
import Definitions.Def_PassivityTorus_power

open Matrix BigOperators
Formal statement
namespace PassivityTorus
theorem passive_iff_symm (q L d : ℕ) [NeZero L] (hL : 3 ≤ L)
    (W : Site q L → Fin q → Matrix (Fin d) (Fin d) ℝ) :
    (∀ v : Site q L → (Fin d → ℝ), power q L d W v = 0)
      ↔ ∀ x a, (W x a)ᵀ = W x a := by sorry
end PassivityTorus
Source
Shape Zero LLC, "Formal Proofs of the C1 Verification Package" (August 2026), §6, Theorem 6.1 (extended to a q-dimensional periodic lattice): https://github.com/ShapeZeroSZ/shape-zero/blob/main/01_source/proofs/ShapeZero_C1_Formal_Proofs.pdf ; corrected in "Errata — C1 Formal Proofs (Sections 3 and 6)", Corrected Theorem 6.1(a): https://github.com/ShapeZeroSZ/shape-zero/blob/main/01_source/proofs/ERRATUM_Theorem_6.1.md
Read-back

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

Setting. Let q,L,dq, L, dq,L,d be natural numbers. Assume L≠0L \neq 0L=0 (a typeclass assumption), and also assume L≥3L \ge 3L≥3 (an explicit hypothesis, which already implies L≠0L \neq 0L=0). A site is a function x:{0,…,q−1}→Z/LZx : \{0,\dots,q-1\} \to \mathbb{Z}/L\mathbb{Z}x:{0,…,q−1}→Z/LZ, so the set of sites is the discrete torus (Z/LZ)q(\mathbb{Z}/L\mathbb{Z})^q(Z/LZ)q, which has LqL^qLq elements. For a site xxx and a direction a∈{0,…,q−1}a \in \{0,\dots,q-1\}a∈{0,…,q−1}, write eae_aea​ for the aaa-th unit vector. The shifted site x±eax \pm e_ax±ea​ is xxx with only its aaa-th coordinate changed, to xa±1x_a \pm 1xa​±1 taken modulo LLL. Addition wraps around: L−1L-1L−1 goes to 000. For −1-1−1, the code adds the residue L−1L-1L−1, so 000 goes to L−1L-1L−1. All other coordinates stay the same. Let

W:(Z/LZ)q×{0,…,q−1}→Rd×dW : (\mathbb{Z}/L\mathbb{Z})^q \times \{0,\dots,q-1\} \to \mathbb{R}^{d\times d}W:(Z/LZ)q×{0,…,q−1}→Rd×d

be an arbitrary assignment of a real d×dd \times dd×d matrix Wx,aW_{x,a}Wx,a​ to every site xxx and direction aaa. No further conditions are placed on WWW.

The quantity PW(v)P_W(v)PW​(v) ("power"). For any function v:(Z/LZ)q→Rdv : (\mathbb{Z}/L\mathbb{Z})^q \to \mathbb{R}^dv:(Z/LZ)q→Rd (arbitrary, and possibly zero), define

PW(v)  =  ∑x∈(Z/LZ)q  ∑a=0q−1v(x)⊤(Wx,a v(x+ea)  −  Wx−ea, a v(x−ea)).P_W(v) \;=\; \sum_{x \in (\mathbb{Z}/L\mathbb{Z})^q} \;\sum_{a=0}^{q-1} v(x)^{\top}\Big( W_{x,a}\, v(x+e_a) \;-\; W_{x-e_a,\,a}\, v(x-e_a) \Big).PW​(v)=x∈(Z/LZ)q∑​a=0∑q−1​v(x)⊤(Wx,a​v(x+ea​)−Wx−ea​,a​v(x−ea​)).

Here u⊤w=∑iuiwiu^\top w = \sum_{i} u_i w_iu⊤w=∑i​ui​wi​ is the ordinary dot product on Rd\mathbb{R}^dRd, and Wx,av(⋅)W_{x,a} v(\cdot)Wx,a​v(⋅) is the ordinary matrix–vector product. The second term uses the matrix at the neighbouring site x−eax-e_ax−ea​, again with direction aaa, not the matrix at xxx.

Assertion. For every q,L,dq, L, dq,L,d and WWW as above (with L≥3L \ge 3L≥3):

(∀ v:(Z/LZ)q→Rd,    PW(v)=0)  ⟺  (∀ x∈(Z/LZ)q, ∀ a∈{0,…,q−1},    Wx,a⊤=Wx,a).\Big(\forall\, v : (\mathbb{Z}/L\mathbb{Z})^q \to \mathbb{R}^d,\;\; P_W(v) = 0\Big) \iff \Big(\forall\, x \in (\mathbb{Z}/L\mathbb{Z})^q,\ \forall\, a \in \{0,\dots,q-1\},\;\; W_{x,a}^{\top} = W_{x,a}\Big).(∀v:(Z/LZ)q→Rd,PW​(v)=0)⟺(∀x∈(Z/LZ)q, ∀a∈{0,…,q−1},Wx,a⊤​=Wx,a​).

In words, PWP_WPW​ vanishes for every vector field vvv exactly when every matrix Wx,aW_{x,a}Wx,a​ is symmetric. The statement is a biconditional, so it claims both directions.

Degenerate cases.

  • If q=0q = 0q=0, there is exactly one site (the empty function) and no directions. The inner sum is then empty, so PW(v)=0P_W(v) = 0PW​(v)=0 for every vvv, and the right-hand side holds vacuously. Both sides are true.
  • If d=0d = 0d=0, all vectors and matrices are empty, so both sides are trivially true.
  • The cases L=1L = 1L=1 and L=2L = 2L=2 (where x+ea=x−eax + e_a = x - e_ax+ea​=x−ea​ or x±ea=xx \pm e_a = xx±ea​=x) are excluded by the hypothesis L≥3L \ge 3L≥3. When L≥3L \ge 3L≥3, the three sites x−eax - e_ax−ea​, xxx, x+eax + e_ax+ea​ are distinct.
Human review
  • Endorsed by Shuze Chen · Sep 24, 2026

    Confirmed by the moderator at approval.

  • Endorsed by ShapeZero · Sep 24, 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 paper
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me