OctoInformBeta

OpenAI preprint, September 2026 · §5. From an ultrapower embedding to a finite contradiction

Lemma 5.3: On a set in the free ultrafilter, for an injective endomorphism with finite-index image

Matched irreducibles have uniform pullback predecessors.

Hover over a blue or green term in the text, or a node in the graph, to find it in the other. Select it to read its declaration both in Lean and in words.

The Setting

A unitary tuple can match finite-dimensional irreducible representations through its quadratic displacement energy on a tensor product. The issue is whether pullback predecessors are available at the same indices for all matched irreducibles.

Let Γ\Gamma and KK be groups, let f:Γ→Kf:\Gamma\to K be a homomorphism, and let θ:Γ→Γ\theta:\Gamma\to\Gamma be an injective endomorphism with finite-index image. Fix t∈Kt\in K, a finite subset S⊆ΓS\subseteq\Gamma and real numbers κ,a\kappa,a. Let BB be a unital complex C*-algebra and let ρ:B→ℬ(E)\rho:B\to\mathcal B(E) be a unital star homomorphism on a complex Hilbert space EE. Inner products are linear in the second variable. Let ω\omega be a free ultrafilter on ℕ\mathbb N, and let un:K→𝒰(B)u_n:K\to\mathcal U(B) be a unitary tuple for each n∈ℕn\in\mathbb N.

Representations and Matching

A finite-dimensional unitary representation π\pi of Γ\Gamma is a homomorphism π:Γ→𝒰(Hπ)\pi:\Gamma\to\mathcal U(H_\pi) on a nonzero finite-dimensional complex Hilbert space HπH_\pi.✦ A finite-dimensional unitary representation π\pi of Γ\Gamma is irreducible if every invariant complex linear subspace of HπH_\pi is either zero or HπH_\pi.✦ Write [π][\pi] for its unitary equivalence class.

Isometric intertwining containment of representations π\pi and τ\tau, written π≺τ\pi\prec\tau, means that there is a complex linear isometry T:Hπ→HτT:H_\pi\to H_\tau satisfying (1): Tπ(g)=τ(g)T(g∈Γ).T\pi(g)=\tau(g)T\qquad(g\in\Gamma).(1)✦ The pullback θ*τ=τ∘θ\theta^*\tau=\tau\circ\theta is the representation on HτH_\tau whose action at gg is τ(θ(g))\tau(\theta(g)).✦

For complex inner product spaces EE and FF, their Hilbert tensor product E⊗FE\otimes F is the completion of the algebraic tensor product in its natural inner product norm.✦ For bounded complex linear operators AA on EE and CC on FF, the tensor product operator A⊗CA\otimes C is the continuous extension of the operator specified in (2): (A⊗C)(x⊗y)=Ax⊗Cy(x∈E,y∈F).(A\otimes C)(x\otimes y)=Ax\otimes Cy \qquad(x\in E,\ y\in F).(2)✦

The matching unitary Un,π(g)U_{n,\pi}(g) at g∈Γg\in\Gamma is the tensor product unitary on E⊗HπE\otimes H_\pi specified in (3); these operators form the matching tuple Un,πU_{n,\pi}: Un,π(g)=ρ(un(f(g)))⊗π(g)on E⊗Hπ.U_{n,\pi}(g)=\rho\bigl(u_n(f(g))\bigr)\otimes\pi(g) \quad\text{on }E\otimes H_\pi.(3) For a family U:Γ→𝒰(H)U:\Gamma\to\mathcal U(H) of unitaries on a Hilbert space HH, the quadratic displacement energy ℰS(U;x)\mathcal E_S(U;x) at x∈Hx\in H is the real number specified in (4): ℰS(U;x)=∑s∈S∥U(s)x−x∥2.\mathcal E_S(U;x)=\sum_{s\in S}\|U(s)x-x\|^2.(4)

The collection ℳn(a)\mathcal M_n(a) consists of the finite-dimensional unitary representations of Γ\Gamma satisfying the strict matching condition (5): π∈ℳn(a)⇔∃x∈E⊗Hπ,∥x∥=1,ℰS(Un,π;x)<a.\pi\in\mathcal M_n(a) \quad\Longleftrightarrow\quad \exists\,x\in E\otimes H_\pi,\qquad \|x\|=1,\qquad \mathcal E_S(U_{n,\pi};x)<a.(5)✦ The second tensor factor in (3) uses π\pi directly; the paper uses the conjugate representation, so a paper representation corresponds here to its dual. Let DnD_n be the set of unitary equivalence classes of finite-dimensional irreducible unitary representations of Γ\Gamma passing the matching test in the paper’s convention.

A Kazhdan gap κ\kappa for S⊆ΓS\subseteq\Gamma is a real number such that (6) holds for every unitary representation VV of Γ\Gamma on a complex Hilbert space HH having no nonzero invariant vector: κ∥ξ∥2≤ℰS(V;ξ)(ξ∈H).\kappa\|\xi\|^2\leq\mathcal E_S(V;\xi) \qquad(\xi\in H).(6)✦

The Claim

Theorem 1. Suppose SS generates Γ\Gamma and has Kazhdan gap κ\kappa, and suppose 0<a<κ0<a<\kappa. Assume f(θ(g))t=tf(g)f(\theta(g))t=tf(g) for every g∈Γg\in\Gamma, un(1)=1u_n(1)=1 for every nn, and lim⁡n→ω∥un(gh)−un(g)un(h)∥=0\lim_{n\to\omega}\|u_n(gh)-u_n(g)u_n(h)\|=0 for every g,h∈Kg,h\in K. Fix a finite-dimensional irreducible unitary representation π0\pi_0 of Γ\Gamma. Then (7) holds: {n∈ℕ:every irreducible π∈ℳn(a) admits an irreducibleτ∈ℳn(a) with π≺θ*τ}∈ω.\left\{n\in\mathbb N: \begin{gathered} \text{every irreducible }\pi\in\mathcal M_n(a)\text{ admits an irreducible}\\ \tau\in\mathcal M_n(a)\text{ with }\pi\prec\theta^*\tau \end{gathered} \right\}\in\omega.(7)✦

In the paper, the result reads: the set of nn for which every [π]∈Dn[\pi]\in D_n has some [π′]∈Dn[\pi']\in D_n satisfying π≺π′∘θ\pi\prec\pi'\circ\theta belongs to ω\omega.

What Is Checked

The statement is over finite-dimensional unitary representations of Γ\Gamma matched in the supplied unital Hilbert realization of BB, under a Kazhdan gap and an asymptotically multiplicative unitary tuple.

For unitaries AA and CC, their tensor product unitary W=A⊗CW=A\otimes C satisfies the adjoint identities (8): W*W=I,WW*=I.W^*W=I,\qquad WW^*=I.(8)✦