OctoInformBeta

OpenAI preprint, September 2026 · §6. Exact transmission on the regular harmonic subspaces

Section 6: regular values, slopes and the ideal pulse: Angular polynomial evolution and its radial synthesis in Cartesian coordinates

Regular values and their logarithmic slopes.

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

Radial evolution of harmonic data changes the coefficients within each angular polynomial space. Its value and velocity are operators on that space, while Cartesian evaluation turns the evolving polynomial into a function of position. The library defines the regular value, its logarithmic slope and the associated radial synthesis.

Let nn be the Cartesian dimension and m=n−1m=n-1 the link dimension. The real polynomial space 𝒫n=ℝ[x1,…,xn]\mathcal P_n=\mathbb R[x_1,\ldots,x_n] consists of polynomials on ℝn\mathbb R^n.✦ The Euclidean polynomial Laplacian Δpol\Delta_{\mathrm{pol}} and the space VℓV_\ell of homogeneous harmonic polynomials of degree ℓ\ell are defined in (1); polynomial evaluation is denoted by P(x)P(x). ΔpolP=∑i=1n∂i2P,Vℓ={P∈𝒫n:P is homogeneous of degree ℓ,ΔpolP=0}.\Delta_{\mathrm{pol}}P=\sum_{i=1}^n\partial_i^2P, \qquad V_\ell=\{P\in\mathcal P_n:P\text{ is homogeneous of degree }\ell, \ \Delta_{\mathrm{pol}}P=0\}.(1)

An orthogonal complex structure JJ on ℝn\mathbb R^n is a matrix in Mn(ℝ)M_n(\mathbb R) satisfying J2=−IJ^2=-I and J𝖳J=IJ^{\mathsf T}J=I.✦ For a real matrix M∈Mn(ℝ)M\in M_n(\mathbb R), the infinitesimal coordinate motion 𝒟M\mathscr D_M is polynomial differentiation along MxMx.✦ The infinitesimal Hopf rotation DJD_J on VℓV_\ell is the bounded operator obtained by restricting the infinitesimal coordinate motion 𝒟J\mathscr D_J to VℓV_\ell, as in (2). 𝒟MP=∑i,j=1nMijxj∂iP,DJ=𝒟J|Vℓ.\mathscr D_M P=\sum_{i,j=1}^n M_{ij}x_j\partial_iP, \qquad D_J=\mathscr D_J\big|_{V_\ell}.(2)

The round eigenvalue in degree ℓ\ell is BℓB_\ell in (3).✦ For profiles J(t)J(t), horizontal metric coefficient c(t)c(t) and squared Hopf stretch q(t)q(t) depending on logarithmic radius tt, the positive angular operator Aℓ(t)A_\ell(t) is the negative Berger Laplacian in (3). Bℓ=ℓ(ℓ+m−1),Aℓ(t)=c(t)−1[BℓI+(1−q(t)−1)DJ(t)2].B_\ell=\ell(\ell+m-1), \qquad A_\ell(t)=c(t)^{-1} \bigl[B_\ell I+(1-q(t)^{-1})D_{J(t)}^2\bigr].(3)

Evolution and Evaluation

Let EE be a real normed space, A(t)A(t) a bounded linear operator on EE and b(t)b(t) a real scalar drift. The first-order block operator KA,b(t)K_{A,b}(t) on E×EE\times E is (4). KA,b(t)(u,v)=(v,A(t)u−b(t)v).K_{A,b}(t)(u,v)=(v,A(t)u-b(t)v).(4)✦

For a bounded operator coefficient C(t)C(t) on a real normed space EE and an initial vector z∈Ez\in E, the Peano-Baker iterates Tj(C,z;t)T_j(C,z;t) are the EE-valued functions given by the successive oriented integrals in (5).✦ For a bounded operator coefficient C(t)C(t) on a real normed space EE and an initial vector z∈Ez\in E, the linear evolution ℱC(t)z\mathcal F_C(t)z is the EE-valued function given by the sum of the Peano-Baker iterates in (5). T0(C,z;t)=z,Tj+1(C,z;t)=∫0tC(s)Tj(C,z;s)ds,ℱC(t)z=∑j=0∞Tj(C,z;t).\begin{aligned} T_0(C,z;t)&=z,& T_{j+1}(C,z;t)&=\int_0^t C(s)T_j(C,z;s)\,ds,\\ \mathcal F_C(t)z&=\sum_{j=0}^{\infty}T_j(C,z;t). \end{aligned}(5)

For a bounded operator coefficient A(t)A(t) on a real normed space EE, a scalar drift b(t)b(t), a real central parameter ℓ\ell and x∈Ex\in E, the regular solution phase ΦA,b,ℓ(x,t)\Phi_{A,b,\ell}(x,t) is the E×EE\times E-valued evolution from the initial phase (x,ℓx)(x,\ell x).✦ The regular solution value yA,b,ℓy_{A,b,\ell} and regular solution velocity vA,b,ℓv_{A,b,\ell} are the EE-valued first and second components, respectively, of the regular solution phase ΦA,b,ℓ\Phi_{A,b,\ell} in (6), where π1\pi_1 and π2\pi_2 denote the coordinate projections from E×EE\times E to EE. ΦA,b,ℓ(x,t)=ℱKA,b(t)(x,ℓx),yA,b,ℓ(x,t)=π1ΦA,b,ℓ(x,t),vA,b,ℓ(x,t)=π2ΦA,b,ℓ(x,t).\begin{aligned} \Phi_{A,b,\ell}(x,t)&=\mathcal F_{K_{A,b}}(t)(x,\ell x),\\ y_{A,b,\ell}(x,t)&=\pi_1\Phi_{A,b,\ell}(x,t),& v_{A,b,\ell}(x,t)&=\pi_2\Phi_{A,b,\ell}(x,t). \end{aligned}(6)

The left composition operator LA(t)L_A(t) acts on bounded endomorphisms of EE by LA(t)T=A(t)∘TL_A(t)T=A(t)\circ T.✦ The regular velocity operator Zℓ(t)Z_\ell(t) on EE is the bounded endomorphism vLA,b,ℓ(I,t)v_{L_A,b,\ell}(I,t).✦ For an endomorphism TT, write T[−1]T^{[-1]} for its inverse when it is invertible and for zero otherwise.

For a polynomial P∈VℓP\in V_\ell and a basis B=(Bi)i∈ℐB=(B_i)_{i\in\mathcal I} of VℓV_\ell, the harmonic polynomial coordinates 𝒄B(P)\mathbf c_B(P) are the coefficients of PP in the basis BB, as in (7). P=∑i∈ℐ𝒄B(P)iBi.P=\sum_{i\in\mathcal I}\mathbf c_B(P)_iB_i.(7)✦

The squared radius r2r^2 and logarithmic radius tt on ℝn\mathbb R^n are the real-valued functions defined by (8); away from the origin, rr is the Euclidean radius and t(x)=log⁡r(x)t(x)=\log r(x). r2(x)=x⋅x,t(x)={12log⁡(r2(x)),x≠0,0,x=0.r^2(x)=x\cdot x, \qquad t(x)=\begin{cases}\tfrac12\log(r^2(x)),&x\ne0,\\0,&x=0.\end{cases}(8)

For a finite family 𝒑=(pi)i∈ℐ\mathbf p=(p_i)_{i\in\mathcal I} of real-valued functions on ℝn\mathbb R^n and a coefficient function 𝒚\mathbf y with values in ℝℐ\mathbb R^{\mathcal I}, the polynomial-normalized coefficient αℓ,i(𝒚;t)\alpha_{\ell,i}(\mathbf y;t) is a real-valued function of tt, and the radial angular synthesis 𝒮ℓ(𝒑,𝒚)\mathcal S_\ell(\mathbf p,\mathbf y) is a real-valued function on ℝn\mathbb R^n, given by (9). αℓ,i(𝒚;t)=e−ℓt𝒚(t)i,𝒮ℓ(𝒑,𝒚)(x)=∑i∈ℐαℓ,i(𝒚;t(x))pi(x).\alpha_{\ell,i}(\mathbf y;t)=e^{-\ell t}\mathbf y(t)_i, \qquad \mathcal S_\ell(\mathbf p,\mathbf y)(x) =\sum_{i\in\mathcal I}\alpha_{\ell,i}(\mathbf y;t(x))p_i(x).(9)

The Definition

Definition 1. Let A(t)A(t) be a bounded operator coefficient on a real normed space EE, let b(t)b(t) be a scalar drift and let ℓ\ell be a real central parameter. The operator-valued regular solution Yℓ(t)Y_\ell(t) is defined by (10). Yℓ(t)=yLA,b,ℓ(I,t).Y_\ell(t)=y_{L_A,b,\ell}(I,t).(10)✦ The regular logarithmic derivative Sℓ(t)S_\ell(t) of the operator-valued regular solution Yℓ(t)Y_\ell(t) on EE is the operator defined by (11). Sℓ(t)=Zℓ(t)Yℓ(t)[−1].S_\ell(t)=Z_\ell(t)Y_\ell(t)^{[-1]}.(11)✦ For a nonnegative integer degree ℓ\ell, angular profiles J,c,qJ,c,q, scalar drift dd and P∈VℓP\in V_\ell, the center-regular angular solution wP(t)w_P(t) is the VℓV_\ell-valued function defined by (12). wP(t)=yAℓ,d,ℓ(P,t).w_P(t)=y_{A_\ell,d,\ell}(P,t).(12)✦ For a nonnegative integer degree ℓ\ell, angular profiles J,c,qJ,c,q, scalar drift dd, a polynomial P∈VℓP\in V_\ell and a basis BB of VℓV_\ell, the center-regular harmonic function uP,Bu_{P,B} is the real-valued function on ℝn\mathbb R^n given by the synthesis in (13). uP,B(x)=𝒮ℓ((ξ↦Bi(ξ))i∈ℐ,t↦𝒄B(wP(t)))(x).u_{P,B}(x)=\mathcal S_\ell \bigl((\xi\mapsto B_i(\xi))_{i\in\mathcal I}, t\mapsto\mathbf c_B(w_P(t))\bigr)(x).(13)✦

In the paper, the result reads: for fixed finite spectral data, radial scale a(t)a(t) and fixed round scale a*a_*, the degree equations and round constants are (14), p=m−1+ma′a,Yℓ″+pYℓ′−AℓYℓ=0,θℓ2+(m−1)θℓ=a*−2Bℓ,θℓ>0,κℓ=a*−2m−1+2θℓ,\begin{aligned} p&=m-1+m\frac{a'}a,& Y_\ell''+pY_\ell'-A_\ell Y_\ell&=0,\\ \theta_\ell^2+(m-1)\theta_\ell&=a_*^{-2}B_\ell, \quad\theta_\ell>0,& \kappa_\ell&=\frac{a_*^{-2}}{m-1+2\theta_\ell}, \end{aligned}(14) the ideal pulse with pulse parameter hh, pulse time τ\tau, scalar coefficients cν(h)c_\nu(h), profiles βν(τ)\beta_\nu(\tau) and structures JνJ_\nu satisfies (15), Hℓ′=RℓHℓ,Hℓ(0)=I,Rℓ=κℓ∑νcν(h)βν(τ)(BℓI/m+DJν2),H_\ell'=R_\ell H_\ell, \qquad H_\ell(0)=I, \qquad R_\ell=\kappa_\ell\sum_\nu c_\nu(h)\beta_\nu(\tau) \bigl(B_\ell I/m+D_{J_\nu}^2\bigr),(15) and the scalar baseline vℓv_\ell on the pulse beginning at tjt_j, together with its pulse integral ϕℓ,j\phi_{\ell,j}, is specified by (16) vℓ′=a(t)−2Bℓ−p(t)vℓ−vℓ2,vℓ(tj)=θℓ,ϕℓ,j=∫pulse jvℓ(t)dt.v_\ell'=a(t)^{-2}B_\ell-p(t)v_\ell-v_\ell^2, \qquad v_\ell(t_j)=\theta_\ell, \qquad \phi_{\ell,j}=\int_{\text{pulse }j}v_\ell(t)\,dt.(16)

What Is Checked

The definitions (10) and (11) are stated over bounded operator coefficients on a real normed space and a scalar drift; the angular specializations (12) and (13) use homogeneous harmonic polynomials, an orthogonal complex structure at each time and a polynomial basis. Harmonic polynomial coordinates give a continuous linear isomorphism with ℝℐ\mathbb R^{\mathcal I}, and Hopf differentiation preserves sums and multiplication by real scalars.