OctoInformBeta

OpenAI preprint, September 2026 · §3. A strict spectral surplus

Section 3: dwell exponents: The fixed Berger parameters in dimension 1616, indexed by homogeneous harmonic polynomials

Dwell eigenvalues and their growth exponents.

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

The angular eigenvalues of a Berger link determine the radial indicial exponents for harmonic functions. The Cartesian dimension here is n=16n=16, and the link dimension is m=n−1=15m=n-1=15. The spectral blocks are numbered by b=0,1,…b=0,1,\ldots, with polynomial degree ℓ=b+2\ell=b+2; a range of LL blocks covers degrees 2,…,L+12,\ldots,L+1.✦

The real polynomial space 𝒫n=ℝ[x1,…,xn]\mathcal P_n=\mathbb R[x_1,\ldots,x_n] consists of polynomials in the Euclidean coordinates.✦ The Euclidean Laplacian Δ0\Delta_0 on 𝒫n\mathcal P_n is the differential operator in (1), and the degree-ℓ\ell harmonic subspace VℓV_\ell of 𝒫n\mathcal P_n is the subspace defined there; write Nℓ=dim⁡ℝVℓN_\ell=\dim_{\mathbb R}V_\ell and equip VℓV_\ell with the Fischer inner product. Δ0P=∑j=1n∂2P∂xj2,Vℓ={P∈𝒫n:P is homogeneous of degree ℓ,Δ0P=0}.\Delta_0 P=\sum_{j=1}^n\frac{\partial^2P}{\partial x_j^2}, \qquad V_\ell=\{P\in\mathcal P_n:P\text{ is homogeneous of degree }\ell, \ \Delta_0P=0\}.(1)

An orthogonal complex structure JJ on ℝn\mathbb R^n is a real matrix satisfying the first two identities in (2).✦ On ℝ2s=ℝs⊕ℝs\mathbb R^{2s}=\mathbb R^s\oplus\mathbb R^s, the standard complex structure J0J_0 is the block matrix in that equation; here s=8s=8.✦ J2=−In,J𝖳J=In,J0=(0−IsIs0).J^2=-I_n,\qquad J^{\mathsf T}J=I_n, \qquad J_0=\begin{pmatrix}0&-I_s\\ I_s&0\end{pmatrix}.(2)

For M∈Mn(ℝ)M\in M_n(\mathbb R), the space of real n×nn\times n matrices, the infinitesimal coordinate motion ℛM\mathcal R_M on 𝒫n\mathcal P_n is the differential operator along the linear vector field MxMx given in (3). The infinitesimal Hopf derivative DJD_J on VℓV_\ell is the restriction in that equation, regarded as a continuous linear operator; its square is DJ2=DJ∘DJD_J^2=D_J\circ D_J. The standard Hopf square is DJ02D_{J_0}^2.✦ ℛMP=∑i,j=1nMijxj∂P∂xi,DJ=ℛJ|Vℓ.\mathcal R_M P=\sum_{i,j=1}^n M_{ij}x_j\frac{\partial P}{\partial x_i}, \qquad D_J=\mathcal R_J\big|_{V_\ell}.(3)

For positive angular parameters a,qa,q, the positive Berger angular operator Aℓ(a,q,J)A_\ell(a,q,J) on VℓV_\ell is given by (4); BℓB_\ell is the round eigenvalue and cc is the angular scale appearing there. Bℓ=ℓ(ℓ+m−1),c=a2q−1/m,Aℓ(a,q,J)=c−1[BℓI+(1−q−1)DJ2].B_\ell=\ell(\ell+m-1),\qquad c=a^2q^{-1/m},\qquad A_\ell(a,q,J)=c^{-1} \bigl[B_\ell I+(1-q^{-1})D_J^2\bigr].(4)

Spectral Coordinates

Let EE be a finite-dimensional real inner product space, with orthonormal basis e=(ei)i=1Ne=(e_i)_{i=1}^N. The orthonormal matrix representation [T]e[T]_e of a continuous linear operator T:E→ET:E\to E in the orthonormal basis ee is the matrix defined in (5); the assignment T↦[T]eT\mapsto[T]_e is linear and continuous.✦ An orthonormal spectral decomposition of a continuous linear operator T:E→ET:E\to E on a finite-dimensional real inner product space EE consists of an orthonormal basis ee of EE and real numbers μi\mu_i satisfying the last two identities in (5).✦ [T]e,ij=⟨ei,Tej⟩,Tei=μiei,[T]e=diag⁡(μ1,…,μN).[T]_{e,ij}=\langle e_i,Te_j\rangle,\qquad Te_i=\mu_i e_i,\qquad [T]_e=\operatorname{diag}(\mu_1,\ldots,\mu_N).(5)

For each ℓ≥2\ell\ge2, choose an orthonormal spectral decomposition (eℓ,i,μℓ,i)i=1Nℓ(e_{\ell,i},\mu_{\ell,i})_{i=1}^{N_\ell} of the standard Hopf square DJ02D_{J_0}^2 on VℓV_\ell.✦ The eigenbasis need not be sorted. Repeated values of μℓ,i\mu_{\ell,i} retain their separate basis indices. The numbers μℓ,i\mu_{\ell,i} are the eigenvalues of the standard Hopf square on VℓV_\ell in the chosen orthonormal eigenbasis.

The fixed Berger parameter q*q_*, metric scale a*a_*, and inverse squared scale β\beta are the constants in (6). q*=101100,a*=99853100000q*1/15,β=a*−2.q_*=\frac{101}{100},\qquad a_*=\sqrt{\frac{99853}{100000}\,q_*^{1/15}},\qquad \beta=a_*^{-2}.(6)

The Definition

Definition 1. For the parameters in (6), let ℓ≥2\ell\ge2 and 1≤i≤Nℓ1\le i\le N_\ell. The dwell eigenvalue λℓ,i\lambda_{\ell,i} of Aℓ(a*,q*,J0)A_\ell(a_*,q_*,J_0) is the real number defined by (7): λℓ,i=β[q*1/15Bℓ+(q*1/15−q*1/15−1)μℓ,i].\lambda_{\ell,i} =\beta\left[q_*^{1/15}B_\ell+ \left(q_*^{1/15}-q_*^{1/15-1}\right)\mu_{\ell,i}\right].(7)✦ The growth exponent dℓ,id_{\ell,i} associated with the dwell eigenvalue λℓ,i\lambda_{\ell,i} is the real number defined by (8): dℓ,i=−7+49+λℓ,i.d_{\ell,i}=-7+\sqrt{49+\lambda_{\ell,i}}.(8)✦

In the paper, the result reads: for a,q>0a,q>0, list the eigenvalues of Aℓ(a,q,J0)A_\ell(a,q,J_0) with real multiplicities as λℓ,1(a,q)≤⋯≤λℓ,Nℓ(a,q)\lambda_{\ell,1}(a,q)\le\cdots\le\lambda_{\ell,N_\ell}(a,q) and define dℓ,i(a,q)≥0d_{\ell,i}(a,q)\ge0 by dℓ,i(a,q)(dℓ,i(a,q)+m−1)=λℓ,i(a,q)d_{\ell,i}(a,q)\bigl(d_{\ell,i}(a,q)+m-1\bigr)=\lambda_{\ell,i}(a,q).

What Is Checked

The definitions are stated over the degree-ℓ\ell polynomial Laplacian kernels in 1616 variables, with their Fischer inner products and a chosen orthonormal spectral decomposition of the standard Hopf square. The standard Hopf square is symmetric for this inner product.✦ Every symmetric continuous linear operator on a finite-dimensional real inner product space has a diagonal matrix in an orthonormal eigenbasis.✦