Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

S2.6, p. 26, and eq. (4), p. 3 — the UNiTE network is equivariant under local unitary changes of the atomic bases

Proved
OrbNetEqui.Equivariance.unite_equivariant

by mikedeng1 · Oct 22, 2026 · Mathlib 0df444a (Lean v4.33.1)

equivariancemachine-learningp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1representation-theory

Consider a molecule with atoms u=1,…,du=1,\dots,du=1,…,d, each with a finite atomic-orbital basis {eu;v}\{e_{u;v}\}{eu;v​}, and an N-body tensor TTT of order N=n+1N=n+1N=n+1 over the concatenated basis. Let KKK be a group with subgroups GuG_uGu​, let πu1:K→U(ιu)\pi^1_u:K\to\mathrm U(\iota_u)πu1​:K→U(ιu​) be the representations on the atomic bases and (UL)L∈Λ(U^L)_{L\in\Lambda}(UL)L∈Λ​ unitary representations on the feature spaces VL\mathbb V^LVL, with finitely many labels. Let FFF be the UNiTE network (Methods 1.1): the diagonal reduction h0h^0h0, then t1t_1t1​ steps of block convolution, message passing and point-wise interaction, then t2t_2t2​ point-wise interaction steps. Assume:

  1. (S35) Du πu2(k)=πu1(k) DuD_u\,\pi^2_u(k)=\pi^1_u(k)\,D_uDu​πu2​(k)=πu1​(k)Du​ for all uuu and k∈Guk\in G_uk∈Gu​;
  2. (S33) in intertwining form, CL1;L2t,c,L(UkL1⊗UkL2)=UkL CL1;L2t,c,LC^{t,c,L}_{L_1;L_2}(U^{L_1}_k\otimes U^{L_2}_k)=U^L_k\,C^{t,c,L}_{L_1;L_2}CL1​;L2​t,c,L​(UkL1​​⊗UkL2​​)=UkL​CL1​;L2​t,c,L​, for every step ttt, every uuu and k∈Guk\in G_uk∈Gu​;
  3. (S32) for the diagonal reduction: ψu((πu1(k))⊗NTu)L,ν=UkL ψu(Tu)L,ν\psi_u\big((\pi^1_u(k))^{\otimes N}T_u\big)_{L,\nu}=U^L_k\,\psi_u(T_u)_{L,\nu}ψu​((πu1​(k))⊗NTu​)L,ν​=UkL​ψu​(Tu​)L,ν​ for all uuu and k∈Guk\in G_uk∈Gu​;
  4. invariant attention: αt(g⋅T,g⋅h)=αt(T,h)\alpha^t(g\cdot T,g\cdot h)=\alpha^t(T,h)αt(g⋅T,g⋅h)=αt(T,h) for every step ttt and every g∈∏uGug\in\prod_uG_ug∈∏u​Gu​;
  5. β>0\beta>0β>0 in every EvNorm, and ϵ>0\epsilon>0ϵ>0.

Then for every g=(gu)ug=(g_u)_ug=(gu​)u​ with gu∈Gug_u\in G_ugu​∈Gu​, every tensor TTT and every atom uuu,

F(g⋅T)u=U(gu)⋅F(T)u,F(g\cdot T)_u=U(g_u)\cdot F(T)_u,F(g⋅T)u​=U(gu​)⋅F(T)u​,

where g⋅Tg\cdot Tg⋅T is the local action (S26), (π1(gu1)⊗⋯⊗π1(guN))Tu⃗(\pi^1(g_{u_1})\otimes\cdots\otimes\pi^1(g_{u_N}))T_{\vec u}(π1(gu1​​)⊗⋯⊗π1(guN​​))Tu​ on each sub-tensor, and U(gu)U(g_u)U(gu​) acts on every channel of label LLL by UguLU^L_{g_u}Ugu​L​.

This is the equivariance claim (4), R⋅F(T[Ψ0])≡F(R⋅T[Ψ0])R\cdot F(T[\Psi_0])\equiv F(R\cdot T[\Psi_0])R⋅F(T[Ψ0​])≡F(R⋅T[Ψ0​]), for the local unitary basis rotations (part (a) of the group GGG of S2.2), which the paper proves in Appendix S2.6.

Formalization Note. U(n)\mathrm U(n)U(n) is an abstract group KKK; scalars are complex. The diagonal reduction ψu\psi_uψu​ is any linear map satisfying (S32), which includes the paper's tabulated one (Corollary S3). The multilayer perceptrons are arbitrary functions, the EvNorm statistics are fixed parameters, the attention weights at each step are an arbitrary function of the step's inputs satisfying the paper's invariance condition, and (S33) is used in intertwining form. Matching weights carry the multiplicity index of (16) and (28) and may depend on the atom and step. Permutations (b) and translations (c) of the symmetry group are not part of the statement.

Preamble
import Mathlib
import Definitions.Def_OrbNetEqui_Equivariance_Basic
Formal statement
namespace OrbNetEqui.Equivariance

open Matrix

theorem unite_equivariant {K : Type*} [Group K] {d n : ℕ} {ι : Fin d → Type*}
    [∀ u, Fintype (ι u)] [∀ u, DecidableEq (ι u)] {Λ : Type*} [Fintype Λ] [DecidableEq Λ]
    {dim : Λ → ℕ} {nmax I J : ℕ} {Kmul : Fin d → Λ → ℕ}
    (S : Fin d → Subgroup K)
    (π₁ : (u : Fin d) → K →* Matrix.unitaryGroup (ι u) ℂ)
    (U : (L : Λ) → K →* Matrix.unitaryGroup (Fin (dim L)) ℂ)
    (D : (u : Fin d) → Matrix (ι u) (MIdx dim Kmul u) ℂ)
    (θ : UniteParams d n ι Λ dim nmax Kmul I J)
    (hD : ∀ u, ∀ k ∈ S u, D u * π₂mat U Kmul u k = (π₁ u k : Matrix (ι u) (ι u) ℂ) * D u)
    (hC : ∀ (t : ℕ) (u : Fin d), ∀ k ∈ S u, ∀ (c : Fin nmax) (L₁ L₂ L : Λ),
      (θ.P t).C c L₁ L₂ L * Matrix.kroneckerMap (· * ·)
          (U L₁ k : Matrix (Fin (dim L₁)) (Fin (dim L₁)) ℂ)
          (U L₂ k : Matrix (Fin (dim L₂)) (Fin (dim L₂)) ℂ) =
        (U L k : Matrix (Fin (dim L)) (Fin (dim L)) ℂ) * (θ.P t).C c L₁ L₂ L)
    (hψ : ∀ u, ∀ k ∈ S u, ∀ (T : (Fin (n + 1) → ι u) → ℂ) (L : Λ) (ν : Fin (θ.μd u L)),
      θ.ψ u (diagAct (π₁ u k : Matrix (ι u) (ι u) ℂ) T) L ν =
        (U L k : Matrix (Fin (dim L)) (Fin (dim L)) ℂ) *ᵥ θ.ψ u T L ν)
    (hα : ∀ (t : ℕ) (g : Fin d → K), (∀ u, g u ∈ S u) →
      ∀ (T : (Fin (n + 1) → Orb ι) → ℂ) (h : Fin d → Feature Λ dim nmax),
        θ.α t (tensorAct π₁ g T) (fun u => featAct U (g u) (h u)) = θ.α t T h)
    (hβ : ∀ (t : ℕ) (L : Λ) (c : Fin nmax), 0 < (θ.P t).βh L c ∧ 0 < (θ.P t).βq L c)
    (hε : 0 < θ.ε)
    (g : Fin d → K) (hg : ∀ u, g u ∈ S u) (T : (Fin (n + 1) → Orb ι) → ℂ) (u : Fin d) :
    unite θ D (tensorAct π₁ g T) u = featAct U (g u) (unite θ D T u) := by sorry

end OrbNetEqui.Equivariance
Source
Qiao et al., Informing Geometric Deep Learning with Electronic Interactions to Accelerate Quantum Chemistry, arXiv:2105.14655v4, p. 26, S2.6 G-equivariance (claim (4), p. 3; definition of G-equivariance, S2.2, p. 23)
Human review
  • Endorsed by Shuze Chen · Oct 22, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 22, 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