OctoInformBeta

OpenAI preprint, September 2026 · §7. Global growth and cone geometry

Section 7: proof of Theorem 1.1: Prescribed transmission histories supply independent harmonic functions with growth order 5000050000

Harmonic growth beyond the Euclidean count.

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

Polynomial growth for a harmonic function is measured using the intrinsic distance of its metric. The construction takes place on ℝn\mathbb R^n with n=16n=16 and link dimension m=n−1=15m=n-1=15; its growth order is k=50000k=50000. A smooth Riemannian metric gg has a smooth, positive definite coefficient matrix G(x)G(x).✦ Its Laplace-Beltrami operator Δg\Delta_g is the trace of the covariant Hessian, with Christoffel symbols determined by GG.✦ For a continuously differentiable curve γ\gamma, its length and the intrinsic distance are defined by (1). Lg(γ)=∫01γ′(s)TG(γ(s))γ′(s)ds,dg(x,y)=infγ∈C1(ℝ,ℝn)γ(0)=x,γ(1)=yLg(γ).L_g(\gamma)=\int_0^1 \sqrt{\gamma'(s)^T G(\gamma(s))\gamma'(s)}\,ds, \qquad d_g(x,y)=\inf_{\substack{\gamma\in C^1(\mathbb R,\mathbb R^n)\\ \gamma(0)=x,\ \gamma(1)=y}}L_g(\gamma).(1) The harmonic growth class Hk(g)H_k(g) consists of the smooth real functions satisfying (2). The numerical Euclidean comparator is Cn,kC_{n,k} in (3). Hk(g)={u∈C∞(ℝn):Δgu=0,∃C≥0∀x,|u(x)|≤C(1+dg(0,x))k}.H_k(g)=\left\{u\in C^\infty(\mathbb R^n): \Delta_g u=0,\quad \exists C\ge0\ \forall x,\quad |u(x)|\le C(1+d_g(0,x))^k\right\}.(2)✦ Cn,k=(n+k−1k)+(n+k−2k−1).C_{n,k}=\binom{n+k-1}{k}+\binom{n+k-2}{k-1}.(3)✦

An orthogonal complex structure JJ on ℝ16\mathbb R^{16} is a real matrix satisfying J2=−IJ^2=-I and JTJ=IJ^TJ=I.✦ The standard structure J0J_0 is the matrix (0−I8I80)\left(\begin{smallmatrix}0&-I_8\\I_8&0\end{smallmatrix}\right) in ℝ8⊕ℝ8\mathbb R^8\oplus\mathbb R^8, with the two blocks identified with the sixteen Cartesian coordinates.✦ For x≠0x\ne0, write r=x⋅xr=\sqrt{x\cdot x} and t=log⁡rt=\log r. The Cartesian Berger tensor determined by profiles c,qc,q and a label J(t)J(t) is (4); its radial and Hopf unit vectors are x/rx/r and J(t)x/rJ(t)x/r. G(0)=I,G(x)=c(t)I+(1−c(t))x⊗xr2+c(t)(q(t)−1)J(t)x⊗J(t)xr2(x≠0).\begin{split} G(0)&=I,\\ G(x)&=c(t)I+(1-c(t))\frac{x\otimes x}{r^2} +c(t)(q(t)-1)\frac{J(t)x\otimes J(t)x}{r^2} \qquad(x\ne0). \end{split}(4)✦

The fixed endpoint parameters q*q_* and a*a_* are the positive numbers specified by (5); the radial profile a(t)a(t) is a different quantity. Let ρ\rho be the fixed smooth transition function used in the construction. The scale decay weight ww, its primitive II and its total I∞I_\infty are given in (6). Choose the scale parameter μ\mu supplied by prescribed-scale attainability. q*=101100,a*2=99853100000q*1/15,a*>0,β=a*−2.q_* =\frac{101}{100},\qquad a_*^2=\frac{99853}{100000}q_*^{1/15},\qquad a_*>0,\qquad \beta=a_*^{-2}.(5) w(t)={0,t≤5/2,ρ(t−5/2)(1+t)−5/4,t>5/2,I(t)=∫2tw(s)ds,I∞=I(4)+4⋅5−1/4,a(t)=e−μI(t),sa(t)=−μw(t),α(t)=a(t)−2,p(t)=14+15sa(t),c(t)=a(t)2q(t)−1/15.\begin{aligned} w(t)&=\begin{cases} 0,&t\le5/2,\\ \rho(t-5/2)(1+t)^{-5/4},&t>5/2, \end{cases} & I(t)&=\int_2^t w(s)\,ds,\\ I_\infty&=I(4)+4\cdot5^{-1/4}, & a(t)&=e^{-\mu I(t)},\\ s_a(t)&=-\mu w(t),& \alpha(t)&=a(t)^{-2},\\ p(t)&=14+15s_a(t),& c(t)&=a(t)^2q(t)^{-1/15}. \end{aligned}(6) Here sas_a is the radial scale logarithmic slope, α\alpha the radial coefficient, and pp the radial drift. The anisotropy logarithmic half slope vv is the scalar function given by v=q′/(2q)v=q'/(2q), and the horizontal logarithmic slope uu is the scalar function given by u=sa−v/15u=s_a-v/15. The inverse squared endpoint scale β\beta is the positive number given by β=a*−2\beta=a_*^{-2}. The horizontal scale coefficient cc is the function given by c(t)=a(t)2q(t)−1/15c(t)=a(t)^2q(t)^{-1/15}. The shifted transition ρ(t−5/2)\rho(t-5/2) is the scale cutoff, equal to zero for t≤5/2t\le5/2 and one for t≥7/2t\ge7/2.

Angular Data and Selected Modes

The homogeneous harmonic polynomial space VlV_l is the subspace of ℝ[x1,…,xn]\mathbb R[x_1,\ldots,x_n] defined by (7), equipped with the Fischer inner product.✦ Its dimension is NlN_l. Polynomial evaluation means substitution of the Cartesian coordinates✦, and a choice of basis identifies VlV_l with its coefficient space.✦ Vl={P:P is homogeneous of degree l,Δ0P=0},Δ0=∑i=1n∂i2,Nl=dim⁡ℝVl.V_l=\left\{P:\ P\text{ is homogeneous of degree }l,\quad \Delta_0P=0\right\},\qquad \Delta_0=\sum_{i=1}^n\partial_i^2,\qquad N_l=\dim_{\mathbb R}V_l.(7) The infinitesimal coordinate motion generated by a matrix MM sends PP to (Mx)⋅∇P(Mx)\cdot\nabla P.✦ The Hopf derivative DJD_J on VlV_l is the operator that sends a polynomial PP to (Jx)⋅∇P(Jx)\cdot\nabla P. The positive angular operator and the round eigenvalue are defined in (8). DJP=(Jx)⋅∇P,Bl=l(l+m−1),Al(t)=c(t)−1[BlI+(1−q(t)−1)DJ(t)2].D_JP=(Jx)\cdot\nabla P,\qquad B_l=l(l+m-1),\qquad A_l(t)=c(t)^{-1}\left[B_lI+(1-q(t)^{-1})D_{J(t)}^2\right].(8) The Euclidean polynomial Laplacian Δ0\Delta_0 on the real polynomial space ℝ[x1,…,xn]\mathbb R[x_1,\ldots,x_n] is the sum of the second derivatives in the Cartesian coordinate directions.✦

Choose an orthonormal eigenbasis (el,i)i=1Nl(e_{l,i})_{i=1}^{N_l} of DJ02D_{J_0}^2 and denote its eigenvalues by ηl,i\eta_{l,i}. The endpoint Berger eigenvalues, their larger indicial roots and the background limiting roots are defined by (9). λl,i=βq*1/15[Bl+(1−q*−1)ηl,i],dl,i=−7+49+λl,i,dl,i(dl,i+m−1)=λl,i,θl=−7+49+βBl.\begin{aligned} \lambda_{l,i} &=\beta q_*^{1/15}\left[B_l+(1-q_*^{-1})\eta_{l,i}\right],\\ d_{l,i}&=-7+\sqrt{49+\lambda_{l,i}},& d_{l,i}(d_{l,i}+m-1)&=\lambda_{l,i},\\ \theta_l&=-7+\sqrt{49+\beta B_l}.&& \end{aligned}(9) There are L=49962+249=50211L=49962+249=50211 spectral blocks.✦ Block index bb corresponds to polynomial degree l=b+2l=b+2, so their degrees are 2,…,L+1=502122,\ldots,L+1=50212.✦ For each such degree, let Sl⊆{1,…,Nl}S_l\subseteq\{1,\ldots,N_l\} be the selected spectral set, with its fixed enumeration, and let Ml=|Sl|M_l=|S_l|. Let ClC_l be the multiplicity count in degree ll; the prescribed counts and selection condition are (10). Ml={Nl,2≤l≤49963,Cl,49964≤l≤50212,∑i∈Sldl,i<kMl.\begin{aligned} M_l&= \begin{cases} N_l,&2\le l\le49963,\\ C_l,&49964\le l\le50212, \end{cases} &\qquad \sum_{i\in S_l}d_{l,i}&<kM_l. \end{aligned}(10) A paired Hopf prefix certificate in degree 49964≤l≤5021249964\le l\le50212 selects prefix indices 0≤h≤bl0\le h\le b_l and their reflections, where blb_l is a nonnegative integer with 2bl<l2b_l<l, and has positive multiplicity ClC_l and nonnegative squared charge total QlchQ_l^{\mathrm{ch}}. The multiplicity prefix ℳ\mathcal M and squared charge prefix 𝒮\mathcal S are defined in (12); the certificate’s exact prefix identities and strict spectral margin are (11). Cl=2(l+7)7ℳ(l,bl),Qlch=2(l+7)7𝒮(l,bl),100000(101l(l+14)Cl−Qlch)<101⋅99853⋅k(k+14)Cl.\begin{split} C_l&=\frac{2(l+7)}7\mathcal M(l,b_l),\qquad Q_l^{\mathrm{ch}}=\frac{2(l+7)}7\mathcal S(l,b_l),\\ 100000\left(101l(l+14)C_l-Q_l^{\mathrm{ch}}\right) &<101\cdot99853\cdot k(k+14)C_l. \end{split}(11) The functions 𝒬j\mathcal Q_j are the squared charge summation primitives. For 0≤j≤60\le j\le6, the coefficients aja_j are the reflected Vandermonde coefficients; ℬ(x,h)=1h!∏v=0h−1(x−v)\mathcal B(x,h)=\frac1{h!}\prod_{v=0}^{h-1}(x-v) is the binomial polynomial, with empty product one. aj(l)=(−1)jℬ(l+6−j,6−j)(6+j6),ℳ(l,b)=∑j=06aj(l)ℬ(b+7,7+j),𝒮(l,b)=∑j=06aj(l)𝒬j(l,b),𝒬j(l,b)=(l−2j)2ℬ(b+7,7+j)+4(7+j)(2j+1−l)ℬ(b+7,8+j)+4(7+j)(8+j)ℬ(b+7,9+j).\begin{split} a_j(l)&=(-1)^j\mathcal B(l+6-j,6-j)\binom{6+j}{6},\\ \mathcal M(l,b)&=\sum_{j=0}^6a_j(l)\mathcal B(b+7,7+j),\\ \mathcal S(l,b)&=\sum_{j=0}^6a_j(l)\mathcal Q_j(l,b),\\ \mathcal Q_j(l,b)&=(l-2j)^2\mathcal B(b+7,7+j)\\ &\quad+4(7+j)(2j+1-l)\mathcal B(b+7,8+j)\\ &\quad+4(7+j)(8+j)\mathcal B(b+7,9+j). \end{split}(12)

Prescribed Transmission Histories

The index j≥0j\ge0 labels period j+1j+1. Its pulse duration τj\tau_j, dwell duration δj\delta_j and starting logarithmic radius TjT_j are (13). τj=(j+1)6,δj=(j+1)7,Tj=10+∑h=0j−1(δh+4τh).\tau_j=(j+1)^6,\qquad \delta_j=(j+1)^7,\qquad T_j=10+\sum_{h=0}^{j-1}(\delta_h+4\tau_h).(13) A smooth chronological control word is an ordered list of N>0N>0 angular slots JiJ_i together with smooth amplitude functions on a finite-dimensional Euclidean parameter space and a positive parameter radius.✦ The admissible amplitude vectors of a smooth chronological control word are the zero vector and the values of its amplitude functions on its closed parameter ball centered at zero.✦ For slot ii, its smooth step is ρ(s−(2i+1))\rho(s-(2i+1))✦ and its pulse bi(s)b_i(s) is the derivative of that step.✦ The packet duration PP of a smooth chronological control word with NN slots is the positive number given by P=2N+1P=2N+1.✦ For period amplitudes sj,is_{j,i}, the unit interval packet scalar and the individual packet stretches are (14). zj(s)=P∑i=0N−1sj,ibi(Ps),qj,i(s)=1+τj−1Psj,ibi(Ps).z_j(s)=P\sum_{i=0}^{N-1}s_{j,i}b_i(Ps),\qquad q_{j,i}(s)=1+\tau_j^{-1}P s_{j,i}b_i(Ps).(14) For a background operator BlIB_lI, the exact Berger coefficient at stretch qq and direction JJ is q1/15BlI+(q1/15−q1/15−1)DJ2q^{1/15}B_lI+(q^{1/15}-q^{1/15-1})D_J^2.✦ A nonlinear angular packet on VlV_l is the operator family obtained by adding to the background operator BlIB_lI the deviations of the exact Berger coefficients from that background, evaluated at the individual stretches in (14).✦ For slot amplitudes sis_i, the control packet scalar at time vv is the sum of the smooth slot pulses bi(v)b_i(v) weighted by sis_i.✦ The nonlinear Berger control word at normalized time ss and time scale TT is the nonlinear angular packet evaluated at the individual stretches 1+T−1Psibi(Ps)1+T^{-1}P s_i b_i(Ps).✦

The smooth transition ramp RR is the function given by R(s)=ρ(2s−1/2)R(s)=\rho(2s-1/2).✦ For a starting index j0j_0, the dwell window wjw_j and the constructed anisotropy are (15); the diagonal background q0q_0 is the anisotropy profile in (15) with all zj=0z_j=0.✦ wj(t)=R(t−Tj−τjτj)−R(t−Tj−2τj−δjτj),q(t)=1+∑j≥j0[τj−1zj(t−Tjτj)+(q*−1)wj(t)].\begin{split} w_j(t)&=R\left(\frac{t-T_j-\tau_j}{\tau_j}\right) -R\left(\frac{t-T_j-2\tau_j-\delta_j}{\tau_j}\right),\\ q(t)&=1+\sum_{j\ge j_0} \left[\tau_j^{-1}z_j\left(\frac{t-T_j}{\tau_j}\right) +(q_*-1)w_j(t)\right]. \end{split}(15) During an active pulse, a selector chooses the only slot whose pulse may be nonzero; its argument is (t−Tj)/τj(t-T_j)/\tau_j. The global label J(t)J(t) is the complex structure of the slot selected during an active pulse, and it is J0J_0 off active pulses.✦ The selector is locally constant except on open gaps where every packet scalar vanishes for every amplitude choice. The infinite angular coefficient is the exact diagonal background plus the sum of active packet increments, multiplied by α(t)\alpha(t); it is Al(t)A_l(t) in (8).✦ The period anisotropy increment at logarithmic radius tt is τj−1zj((t−Tj)/τj)+(q*−1)wj(t)\tau_j^{-1}z_j((t-T_j)/\tau_j)+(q_*-1)w_j(t), the sum of its pulse and dwell contributions.✦

A prescribed transmission history FF is an amplitude history with admissible values, zero for j<j0j<j_0, with 1/2≤qj,i≤3/21/2\le q_{j,i}\le3/2 in each active packet and 1/2≤q(t)≤21/2\le q(t)\le2. Its radial, horizontal and Hopf Ricci numerators, respectively, are nonnegative at every tt; these are the quantities in (16). ℛr=−[15(u+u′+u2)+v+v′+2uv+v2],ℛh=16−2q−c[u′+(14+15u+v)(1+u)],ℛv=14q2−cq[u′+v′+(14+15u+v)(1+u+v)].\begin{split} \mathcal R_r&=-\left[15(u+u'+u^2)+v+v'+2uv+v^2\right],\\ \mathcal R_h&=16-2q-c\left[u'+(14+15u+v)(1+u)\right],\\ \mathcal R_v&=14q^2-cq\left[u'+v'+(14+15u+v)(1+u+v)\right]. \end{split}(16)

The regular value operator Φl(t)\Phi_l(t) on VlV_l is the operator solution of (17).✦ The regular phase of an initial polynomial P∈VlP\in V_l is the pair (Φl(t)P,Φl′(t)P)(\Phi_l(t)P,\Phi_l'(t)P)✦; the first-order block operator on Vl⊕VlV_l\oplus V_l sends (y,v)(y,v) to (v,Al(t)y−p(t)v)(v,A_l(t)y-p(t)v).✦ Φl″+pΦl′−AlΦl=0,Φl(0)=I,Φl′(0)=lI.\Phi_l''+p\Phi_l'-A_l\Phi_l=0,\qquad \Phi_l(0)=I,\qquad \Phi_l'(0)=lI.(17) For scalar coefficients, the displacement and velocity propagators are the value solutions with initial data (1,0)(1,0) and (0,1)(0,1). On the interval from Tj+τjT_j+\tau_j to Tj+1T_{j+1}, the diagonal leg coefficient on axis ii is (18).✦ Let Uj,lU_{j,l} and Vj,lV_{j,l} be the diagonal matrices of its scalar displacement and velocity propagators across that interval, with drift pp. Λl,i(t)=α(t)[q0(t)1/15Bl+(q0(t)1/15−q0(t)1/15−1)ηl,i].\Lambda_{l,i}(t)=\alpha(t) \left[q_0(t)^{1/15}B_l+ \bigl(q_0(t)^{1/15}-q_0(t)^{1/15-1}\bigr)\eta_{l,i}\right].(18) The period Riccati baseline ζj,l(t)\zeta_{j,l}(t) is y′/yy'/y at elapsed time max⁡(0,t−Tj)\max(0,t-T_j), where yy solves the scalar equation with potential α(Tj+s)Bl\alpha(T_j+s)B_l, drift p(Tj+s)p(T_j+s) and initial data (1,θl)(1,\theta_l).✦ The post-pulse normalization Hj,lH_{j,l} on VlV_l is the diagonal operator given by Hj,l=Uj,l+θlVj,lH_{j,l}=U_{j,l}+\theta_lV_{j,l}.✦

For a time-dependent operator KK on a real normed space EE and an initial vector xx, the Peano-Baker iterates start with the constant function xx; each successive iterate at tt is the oriented integral from zero to tt of K(s)K(s) applied to the preceding iterate at ss.✦ The linear evolution from xx for an operator coefficient KK is the sum of its Peano-Baker iterates over all nonnegative iteration indices.✦ For an operator coefficient AA on EE, the left composition operator at tt sends a bounded endomorphism TT of EE to A(t)∘TA(t)\circ T.✦ For scalar potential AA and drift pp, the scalar solution phase with initial data (a,b)(a,b) is the linear evolution from value aa and velocity bb for the first-order system y′=vy'=v, v′=Ay−pvv'=Ay-pv.✦

Let πl\pi_l be the permutation that cycles the selected coordinates in their fixed enumeration and fixes the others.✦ The signed cycle Pˆl\widehat P_l of the coordinate permutation πl\pi_l is the determinant-one matrix obtained by placing the sign correction on the first selected input axis.✦ For every j≥j0j\ge j_0, the prescribed transmission equality is (19), expressed in the basis (el,i)(e_{l,i}). Here εj,l∈(1/2,3/2)\varepsilon_{j,l}\in(1/2,3/2) is the period correction; set it equal to one when j<j0j<j_0.✦ [Φl(Tj+1)Φl(Tj)−1]=fj,lHj,lPˆl,fj,l=exp⁡(∫TjTj+τjζj,l(t)dt)εj,l.\left[\Phi_l(T_{j+1})\Phi_l(T_j)^{-1}\right] =f_{j,l}H_{j,l}\widehat P_l,\qquad f_{j,l}=\exp\left(\int_{T_j}^{T_j+\tau_j}\zeta_{j,l}(t)\,dt\right) \varepsilon_{j,l}.(19) The virtual transfer Wj,lW_{j,l} on VlV_l is the operator represented by fj,lHj,lPˆlf_{j,l}H_{j,l}\widehat P_l.✦ For x∈Vlx\in V_l, the chronological orbit of the virtual transfers starts at xx and successively applies W0,l,W1,l,…W_{0,l},W_{1,l},\ldots✦; denote the resulting linear operator on VlV_l at index j0j_0 by QlQ_l.✦ For i∈Sli\in S_l, the selected initial polynomial Pl,i∈VlP_{l,i}\in V_l and its global radial angular synthesis ul,i:ℝn→ℝu_{l,i}:\mathbb R^n\to\mathbb R are the polynomial and function given by (20). Pl,i=Φl(Tj0)−1Qlel,i,i∈Sl,ul,i(x)=∑h=1Nle−lt[Φl(t)Pl,i]hel,h(x),t=log⁡r,x≠0,ul,i(0)=0.\begin{split} P_{l,i}&=\Phi_l(T_{j_0})^{-1}Q_le_{l,i},\qquad i\in S_l,\\ u_{l,i}(x)&=\sum_{h=1}^{N_l} e^{-l t}\,[\Phi_l(t)P_{l,i}]_h\,e_{l,h}(x),\qquad t=\log r,\quad x\ne0,\\ u_{l,i}(0)&=0. \end{split}(20) The bracket denotes polynomial coordinates in the chosen basis. The period transmission factor fj,lf_{j,l} is the product of the correction εj,l\varepsilon_{j,l} and the exponential of the integral of ζj,l\zeta_{j,l} over the pulse interval [Tj,Tj+τj][T_j,T_j+\tau_j].✦

The initial data pullback operator on VlV_l first applies the chronological orbit operator QlQ_l and then the inverse regular value operator at Tj0T_{j_0}.✦

The Claim

Theorem 1. Let EE be a finite-dimensional real inner-product space. Let A(t)A(t) be a continuous self-adjoint positive semidefinite operator family on EE, and let p0(t)p_0(t) be a continuous real drift. Assume ∥A(t)∥≤MA\|A(t)\|\le M_A and |p0(t)|≤Mp|p_0(t)|\le M_p for all tt, where MA,Mp≥0M_A,M_p\ge0. Suppose l0>0l_0>0 and B0∈ℝB_0\in\mathbb R satisfy A(t)=l0(l0+B0)IA(t)=l_0(l_0+B_0)I and p0(t)=B0p_0(t)=B_0 for t≤0t\le0. Let Φ\Phi solve Φ″+p0Φ′−AΦ=0\Phi''+p_0\Phi'-A\Phi=0 with initial data (I,l0I)(I,l_0I). For σ∈ℝ\sigma\in\mathbb R and x≠0x\ne0 in EE, the normalized solution YσY_\sigma is the EE-valued function given by Yσ(t;x)=Φ(t)Φ(σ)−1xY_\sigma(t;x)=\Phi(t)\Phi(\sigma)^{-1}x. Suppose a sequence tj>0t_j>0 tends to infinity and satisfies (21) for real numbers 0≤d<γ0\le d<\gamma. log⁡∥Yσ(tj;x)∥tj→d,tj+1−tjtj→0.\frac{\log\|Y_\sigma(t_j;x)\|}{t_j}\longrightarrow d, \qquad \frac{t_{j+1}-t_j}{t_j}\longrightarrow0.(21) Then there is R>0R>0 for which the radial estimate (22) holds. ∥Yσ(log⁡r;x)∥≤rγ(r≥R).\|Y_\sigma(\log r;x)\|\le r^\gamma\qquad(r\ge R).(22)✦ For every prescribed transmission history FF with targets Pˆl\widehat P_l, let gFg_F be its metric (4) and let ℐ={(l,i):2≤l≤50212,i∈Sl}\mathcal I=\{(l,i):2\le l\le50212,\ i\in S_l\}. The selected functions in (20) satisfy (23), and their independence and cardinality are (24) and (25). ul,i∈H50000(gF)((l,i)∈ℐ).u_{l,i}\in H_{50000}(g_F)\qquad((l,i)\in\mathcal I).(23)✦ (ul,i)(l,i)∈ℐ is linearly independent over ℝ.(u_{l,i})_{(l,i)\in\mathcal I} \text{ is linearly independent over }\mathbb R.(24)✦ C16,50000<|ℐ|=∑l=250212Ml.C_{16,50000}<|\mathcal I|=\sum_{l=2}^{50212}M_l.(25)✦

In the paper, the result reads: for the center-regular solution YY chosen on each selected axis and cycle sampling times tJt_J, exact cycles give log⁡∥Y(tJ)∥/tJ→Ml−1∑i∈Sldl,i<k\log\|Y(t_J)\|/t_J\to M_l^{-1} \sum_{i\in S_l}d_{l,i}<k, uniform Riccati bounds and (tj+1−tj)/tj→0(t_{j+1}-t_j)/t_j\to0 give the same upper exponent at every real logarithmic radius, the uniform angular bound and compact absorption give, for a constant CuC_u, the estimate |u(x)|≤Cu(1+dg(0,x))k|u(x)|\le C_u(1+d_g(0,x))^k at every point, the selected functions are independent with hk=dim⁡Hk(g)h_k=\dim H_k(g) satisfying hk≥∑lMl>∑l=0kNlh_k\ge\sum_lM_l>\sum_{l=0}^kN_l, and the volume ratio tends to a*ma_*^m.

What Is Checked

The radial estimate is stated over bounded continuous symmetric nonnegative operator coefficients and a sphere-normalized center-regular solution with the sampling hypotheses in (21). The harmonic family statements are over the prescribed history FF, its Cartesian Berger metric and selected polynomial data. The angular and radial pulse data use exponent 1/151/15, limiting drift 1414, positive drift threshold 1010 and radial coefficient bound 22; their Hopf squares have quadratic forms between −Bl∥P∥2-B_l\|P\|^2 and zero, and their roots satisfy βBl=θl(θl+14)\beta B_l=\theta_l(\theta_l+14) with θl>0\theta_l>0.✦

The standard Hopf square is symmetric for the Fischer inner product✦, and symmetric operators on finite-dimensional real inner-product spaces have orthonormal spectral decompositions whose coordinate matrices are their eigenvalue diagonals. Prescribed-scale attainability gives 0<μ≤1/40<\mu\le1/4, e−μI∞=a*e^{-\mu I_\infty}=a_* and a(t)→a*a(t)\to a_*.✦ The radial coefficient α\alpha is smooth✦, satisfies 0≤α≤20\le\alpha\le2, equals one for t≤0t\le0✦ and tends to β\beta.✦ The drift is continuous, equals 1414 for t≤0t\le0✦ and tends to 1414✦, with 10≤p(t)≤1410\le p(t)\le14. The constructed radial history data consist of these radial profiles, the endpoint anisotropy q*≥1q_*\ge1, the positive degrees ll with Bl=l(l+14)B_l=l(l+14), and the symmetric background and controlled Hopf squares DJ02D_{J_0}^2 and DJ2D_J^2.

For every prescribed history, cc and qq are smooth and positive, with c(t)=q(t)=1c(t)=q(t)=1 for t≤0t\le0. Near each logarithmic radius, either JJ is constant or qq is identically one on a neighborhood.✦ The Cartesian Berger tensor has smooth entries✦ and is positive definite at every point.✦