OctoInformBeta

OpenAI preprint, September 2026 · §4. Geodesic passages and the index form

Lemma 4.2 [Endpoint correction]: The coordinate identity holds in either endpoint frame and gives a scalar correction for a transverse endpoint field

The neck index includes a Christoffel endpoint term.

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

Consider a variation field on a finite geodesic segment in a neck, with unrestricted endpoint values. Write (t,x,y)(t,x,y) for neck coordinates and identify a tangent vector with its axial and transverse parts U=(Ut,U⟂)U=(U_t,U_\perp) in ℝ×ℝ2\mathbb R\times\mathbb R^2.

A smooth Riemannian metric gg on a smooth manifold is a smooth field of positive definite symmetric inner products on its tangent spaces.✦ On a vector space, a smooth field qq of symmetric positive definite continuous bilinear forms with bounded unit balls gives the metric whose inner product at xx is q(x)q(x).✦ The Cartesian bilinear form ⟨⋅,⋅⟩\langle\cdot,\cdot\rangle on a finite coordinate space is its Euclidean inner product.✦ For a linear map LL between vector spaces and a bilinear form qq on its target, the pullback L*qL^*q is the bilinear form on its domain given by (u,w)↦q(Lu,Lw)(u,w)\mapsto q(Lu,Lw).✦

A smooth moving frame K(x)K(x) on a finite coordinate space, consisting of linear automorphisms of that space, defines the metric gx(u,w)=⟨K(x)u,K(x)w⟩g_x(u,w)=\langle K(x)u,K(x)w\rangle.✦ For smooth nonvanishing radii r,ρr,\rho and smooth shear cc, the triangular frame and its inverse are given by (1).✦ The triangular frame metric gr,ρ,cg_{r,\rho,c} is the Riemannian metric determined by the frame Kr,ρ,cK_{r,\rho,c}.✦ Kr,ρ,c(u)=(u0,r(u1−cu2),ρu2),Kr,ρ,c−1(w)=(w0,w1/r+cw2/ρ,w2/ρ).\begin{aligned} K_{r,\rho,c}(u)&=(u_0,r(u_1-cu_2),\rho u_2),\\ K_{r,\rho,c}^{-1}(w)&=(w_0,w_1/r+cw_2/\rho,w_2/\rho). \end{aligned}(1) For coefficients depending only on tt, the cylinder metric in coordinates has line element (2). gr,ρ,c=dt2+r(t)2(dx−c(t)dy)2+ρ(t)2dy2.g_{r,\rho,c}=dt^2+r(t)^2\bigl(dx-c(t)\,dy\bigr)^2+\rho(t)^2dy^2.(2)✦

Let θ\theta denote the standard smooth transition from zero to one. The one-sided cutoff aa, its integrated cutoff 𝒜\mathcal A, and the shear transition χ\chi are the functions in (3). a(t)={e−1/(2/3−t),t<2/3,0,t≥2/3,𝒜(t)=∫0ta(u)du,χ(t)=θ(6(t−5/12)).\begin{aligned} a(t)&=\begin{cases}e^{-1/(2/3-t)},&t<2/3,\\0,&t\geq2/3,\end{cases} &\mathcal A(t)&=\int_0^t a(u)\,du,\\ \chi(t)&=\theta\bigl(6(t-5/12)\bigr). \end{aligned}(3) Let δ\delta be the neck parameter. The diagonal coefficients d1,d2d_1,d_2 and shear σ\sigma determine the transverse cometric PP by (4). d1(δ,t)=1+δ𝒜(t),d2(δ,t)=1+δ𝒜(1−t),σ(δ,t)=δ2χ(t),P(δ,t)=(d1+d2σ2d2σd2σd2).\begin{aligned} d_1(\delta,t)&=1+\delta\mathcal A(t), &d_2(\delta,t)&=1+\delta\mathcal A(1-t), &\sigma(\delta,t)&=\delta^2\chi(t),\\ P(\delta,t)&= \begin{pmatrix}d_1+d_2\sigma^2&d_2\sigma\\d_2\sigma&d_2\end{pmatrix}. \end{aligned}(4) The collar denominator dd, terminal decay rate rate\operatorname{rate}, logarithmic decay kk, warping function ℓδ\ell_\delta, and collar transition λ\lambda are real-valued functions of their displayed arguments defined in (5). d(δ,t)=1+δ𝒜(t),λ(t)=θ(2t+3/2),rate⁡(δ,t)=δa(t)2d(δ,t),k(δ,t)=1−λ(t)+λ(t)rate⁡(δ,t),ℓδ(t)=exp⁡(−∫0tk(δ,u)du).\begin{aligned} d(\delta,t)&=1+\delta\mathcal A(t), &\lambda(t)&=\theta(2t+3/2),\\ \operatorname{rate}(\delta,t)&=\frac{\delta a(t)}{2d(\delta,t)}, &k(\delta,t)&=1-\lambda(t)+\lambda(t)\operatorname{rate}(\delta,t),\\ \ell_\delta(t)&=\exp\left(-\int_0^t k(\delta,u)\,du\right). \end{aligned}(5)

For 0<δ≤1/20<\delta\leq1/2, the normalized neck metric g0g_0 is the cylinder metric (2) with r(t)=ℓδ(t)r(t)=\ell_\delta(t), ρ(t)=ℓδ(1−t)\rho(t)=\ell_\delta(1-t) and c(t)=σ(δ,t)c(t)=\sigma(\delta,t).✦ These radii are smooth and positive. The right collar linear map FδF_\delta and coordinate map fδf_\delta are specified in (6). Fδ(y)=(−y0,δ2y1+y2,y1),Fδ−1(x)=(−x0,x2,x1−δ2x2),fδ(y)=Fδ(y)+(1,0,0).\begin{aligned} F_\delta(y)&=(-y_0,\delta^2y_1+y_2,y_1), &F_\delta^{-1}(x)&=(-x_0,x_2,x_1-\delta^2x_2),\\ f_\delta(y)&=F_\delta(y)+(1,0,0). \end{aligned}(6) The right-framed neck metric g1g_1 is the pullback metric fδ*g0f_\delta^*g_0.✦ For a smooth map ff with invertible differential, the pullback Riemannian metric has tangent inner product (f*g)x(u,w)=gf(x)(dfxu,dfxw)(f^*g)_x(u,w)=g_{f(x)}(df_xu,df_xw).✦ The endpoint-framed neck metric gεg_\varepsilon is the Riemannian metric equal to g0g_0 or g1g_1 according as ε=0\varepsilon=0 or 11.✦

At inward endpoint height τ\tau, the terminal cometric PendP_{\mathrm{end}} and the endpoint coordinate cometrics are given by (7), where C2C_2 is the endpoint frame matrix. Their inverse matrices Qε,δQ_{\varepsilon,\delta} are the transverse metric matrices. C2(δ)=(011−δ2),Pend(δ,τ)=C2(δ)P(δ,1−τ)C2(δ)𝖳,P0,δ(τ)=P(δ,τ),P1,δ(τ)=Pend(δ,τ),Qε,δ(τ)=Pε,δ(τ)−1.\begin{aligned} C_2(\delta)&=\begin{pmatrix}0&1\\1&-\delta^2\end{pmatrix}, &P_{\mathrm{end}}(\delta,\tau) &=C_2(\delta)P(\delta,1-\tau)C_2(\delta)^{\mathsf T},\\ P_{0,\delta}(\tau)&=P(\delta,\tau), &P_{1,\delta}(\tau)&=P_{\mathrm{end}}(\delta,\tau),\\ Q_{\varepsilon,\delta}(\tau)&=P_{\varepsilon,\delta}(\tau)^{-1}. \end{aligned}(7)

Index Densities and Boundary Data

For a smooth metric gg and a coordinate chart cc, the metric matrix (gij)(g_{ij}) consists of the inner products of the vectors d(c−1)eid(c^{-1})e_i, where eie_i is the iith standard basis vector.✦ The coordinate derivative ∂if\partial_i f is Df[ei]Df[e_i].✦ The Christoffel symbols and curvature components are defined by (8), with (gij)=(gij)−1(g^{ij})=(g_{ij})^{-1} and all indices ranging over the three coordinates. Γjki=12∑lgil(∂jglk+∂kglj−∂lgjk),Rlkij=∂iΓjkl−∂jΓikl+∑m(ΓimlΓjkm−ΓjmlΓikm).\begin{aligned} \Gamma^i_{jk} &=\tfrac12\sum_l g^{il} (\partial_jg_{lk}+\partial_kg_{lj}-\partial_lg_{jk}),\\ R^l{}_{kij} &=\partial_i\Gamma^l_{jk}-\partial_j\Gamma^l_{ik} +\sum_m\bigl(\Gamma^l_{im}\Gamma^m_{jk} -\Gamma^l_{jm}\Gamma^m_{ik}\bigr). \end{aligned}(8) The connection term is Γ(u,w)i=∑j,kΓjkiujwk\Gamma(u,w)^i=\sum_{j,k}\Gamma^i_{jk}u^jw^k✦, and the curvature operator has components (R(u,w)z)l=∑k,i,jRlkijzkuiwj(R(u,w)z)^l=\sum_{k,i,j}R^l{}_{kij}z^ku^iw^j.✦

A tangent field JJ along a curve γ\gamma has field coordinates w(s)=dcγ(s)J(s)w(s)=dc_{\gamma(s)}J(s) in cc.✦ Put x=c∘γx=c\circ\gamma. The covariant derivative DsJD_sJ is obtained from the coordinate expression w′+Γ(x′,w)w'+\Gamma(x',w) by applying the inverse chart differential.✦ The curvature term R(J,γ̇)γ̇R(J,\dot\gamma)\dot\gamma is obtained by the same transport of the coordinate curvature operator.✦ A geodesic γ\gamma for a smooth metric gg on a set of times UU is a smooth curve on UU whose covariant acceleration vanishes there in every chart of the atlas.✦ For tangent fields J,HJ,H along a curve γ\gamma in a Riemannian manifold with metric gg, the intrinsic index density ig(J,H;s)i_g(J,H;s) is the scalar function of time defined in (9)✦, and the intrinsic index Ig(J;[a,b])I_g(J;[a,b]) is the integral specified there on the oriented interval [a,b][a,b]. ig(J,H;s)=g(DsJ,DsH)−g(R(J,γ̇)γ̇,H),Ig(J;[a,b])=∫abig(J,J;s)ds.\begin{aligned} i_g(J,H;s)&=g(D_sJ,D_sH) -g\bigl(R(J,\dot\gamma)\dot\gamma,H\bigr),\\ I_g(J;[a,b])&=\int_a^b i_g(J,J;s)\,ds. \end{aligned}(9)

For real 2×22\times2 matrices D,QD,Q, transverse vectors u,zu,z and split vectors U,VU,V, the matrix bilinear pairing ℬD\mathcal B_D is the bilinear form on transverse vectors given in (10)✦, and the split neck metric gQg_Q is the bilinear form on split vectors given there.✦ The axial and transverse splitting of x∈ℝ3x\in\mathbb R^3 is (x0,(x1,x2))(x_0,(x_1,x_2)).✦ ℬD(u,z)=u𝖳Dz,gQ(U,V)=UtVt+ℬQ(U⟂,V⟂).\mathcal B_D(u,z)=u^{\mathsf T}Dz, \qquad g_Q(U,V)=U_tV_t+\mathcal B_Q(U_\perp,V_\perp).(10) Let P,Q,H,KP,Q,H,K be transverse matrices. In height-dependent metric data, HH and KK denote the first and second height derivatives of QQ. The neck Christoffel operator Γ\Gamma and its axial derivative Γt\Gamma_t are defined by (11); subscripts tt and ⟂\perp denote axial and transverse components. ΓP,H(U,V)=(−12ℬH(U⟂,V⟂),12(UtPHV⟂+VtPHU⟂)),(Γt)P,H,K(U,V)=(−12ℬK(U⟂,V⟂),12(Ut(PK−PHPH)V⟂+Vt(PK−PHPH)U⟂)).\begin{aligned} \Gamma_{P,H}(U,V) &=\left(-\tfrac12\mathcal B_H(U_\perp,V_\perp), \tfrac12(U_tPHV_\perp+V_tPHU_\perp)\right),\\ (\Gamma_t)_{P,H,K}(U,V) &=\left(-\tfrac12\mathcal B_K(U_\perp,V_\perp), \tfrac12\bigl(U_t(PK-PHPH)V_\perp +V_t(PK-PHPH)U_\perp\bigr)\right). \end{aligned}(11) The coordinate curvature tensor RP,H,KR_{P,H,K} is the trilinear operator in (12), where the operators on the right have the parameters in (11). RP,H,K(U,V)W=UtΓt(V,W)−VtΓt(U,W)+Γ(U,Γ(V,W))−Γ(V,Γ(U,W)).R_{P,H,K}(U,V)W =U_t\Gamma_t(V,W)-V_t\Gamma_t(U,W) +\Gamma(U,\Gamma(V,W))-\Gamma(V,\Gamma(U,W)).(12)✦

The coordinate index density iQi_Q, coordinate energy density ee, and boundary correction BHB_H of split vectors T=(v,u)T=(v,u), V=(η,z)V=(\eta,z) and V̇=(ξ,ζ)\dot V=(\xi,\zeta) are the scalar functions in (13). iQ=gQ(V̇+ΓP,H(T,V),V̇+ΓP,H(T,V))−gQ(V,RP,H,K(V,T)T),eQ,H,K(T,V,V̇)=ξ2+ℬQ(ζ,ζ)+2ηℬH(ζ,u)+12η2ℬK(u,u),BH(T,V)=−12vℬH(z,z)+ηℬH(u,z).\begin{aligned} i_Q&=g_Q\bigl(\dot V+\Gamma_{P,H}(T,V), \dot V+\Gamma_{P,H}(T,V)\bigr) -g_Q\bigl(V,R_{P,H,K}(V,T)T\bigr),\\ e_{Q,H,K}(T,V,\dot V) &=\xi^2+\mathcal B_Q(\zeta,\zeta) +2\eta\mathcal B_H(\zeta,u) +\tfrac12\eta^2\mathcal B_K(u,u),\\ B_H(T,V)&=-\tfrac12v\mathcal B_H(z,z)+\eta\mathcal B_H(u,z). \end{aligned}(13) The boundary derivative BP,H,K♯(T,V,V̇)B^\sharp_{P,H,K}(T,V,\dot V) is the scalar function of transverse matrices P,H,KP,H,K and split vectors T,V,V̇T,V,\dot V defined by (14). BP,H,K♯(T,V,V̇)=−14ℬH(u,u)ℬH(z,z)−vℬH(ζ,z)−12v2ℬK(z,z)+ξℬH(z,u)+ηℬH(ζ,u)+ηvℬK(z,u)−ηvℬHPH(z,u).\begin{aligned} B^\sharp_{P,H,K}(T,V,\dot V) ={}&-\tfrac14\mathcal B_H(u,u)\mathcal B_H(z,z) -v\mathcal B_H(\zeta,z)-\tfrac12v^2\mathcal B_K(z,z)\\ &+\xi\mathcal B_H(z,u)+\eta\mathcal B_H(\zeta,u) +\eta v\mathcal B_K(z,u)-\eta v\mathcal B_{HPH}(z,u). \end{aligned}(14)✦ For height-dependent data evaluated at t(s)t(s), let E[a,b]E[a,b] be the coordinate energy ∫abeQ,H,K(T,V,V̇)ds\int_a^b e_{Q,H,K}(T,V,\dot V)\,ds, and put B(s)=BH(t(s))(T(s),V(s))B(s)=B_{H(t(s))}(T(s),V(s)). For a curve and field in the endpoint frame, write τ(s)=γ0(s)\tau(s)=\gamma_0(s) and use their axial and transverse splittings T(s)=split⁡(γ′(s))T(s)=\operatorname{split}(\gamma'(s)), V(s)=split⁡(J(s))V(s)=\operatorname{split}(J(s)) and V̇(s)=split⁡(J′(s))\dot V(s)=\operatorname{split}(J'(s)). Denote their coordinate energy and boundary correction by Eε,δ[a,b]E_{\varepsilon,\delta}[a,b] and Bε,δ(s)B_{\varepsilon,\delta}(s), using Q=Qε,δQ=Q_{\varepsilon,\delta}, H=Qε,δ′H=Q'_{\varepsilon,\delta} and K=Qε,δ″K=Q''_{\varepsilon,\delta} at τ(s)\tau(s).

The Claim

Theorem 1 (Endpoint correction). Let P,Q,H,KP,Q,H,K be symmetric real 2×22\times2 matrices with QP=IdQP=\mathrm{Id}. For split vectors T,V,V̇T,V,\dot V, the pointwise identity is (15). iQ=eQ,H,K(T,V,V̇)−BP,H,K♯(T,V,V̇).i_Q=e_{Q,H,K}(T,V,\dot V)-B^\sharp_{P,H,K}(T,V,\dot V).(15)✦

Let P,Q,H,KP,Q,H,K now be matrix-valued functions of height. On the closed interval with endpoints a,ba,b, suppose their compositions with tt and the functions T=(v,u)T=(v,u), V=(η,z)V=(\eta,z) and V̇=(ξ,ζ)\dot V=(\xi,\zeta) are continuous. Assume the four matrices are symmetric and QP=IdQP=\mathrm{Id} along the interval, and that the ordinary derivatives satisfy (16) at every time in it. t′=v,v′=12ℬH(t)(u,u),u′=−vP(t)H(t)u,H′(t(s))=K(t(s)),η′=ξ,z′=ζ.\begin{aligned} t'&=v,&v'&=\tfrac12\mathcal B_{H(t)}(u,u), &u'&=-vP(t)H(t)u,\\ H'(t(s))&=K(t(s)),&\eta'&=\xi,&z'&=\zeta. \end{aligned}(16) Their coordinate index is given by (17). ∫abiQ(s)ds=E[a,b]−(B(b)−B(a)).\int_a^b i_Q(s)\,ds=E[a,b]-\bigl(B(b)-B(a)\bigr).(17)✦

For real transverse matrices P,QP,Q with QP=IdQP=\mathrm{Id}, any transverse matrix HH and split vectors T,VT,V, the Christoffel expression for the correction is (18). BH(T,V)=gQ(T,ΓP,H(V,V)).B_H(T,V)=g_Q\bigl(T,\Gamma_{P,H}(V,V)\bigr).(18)✦ For real κ,v\kappa,v and transverse vectors u,zu,z, its value for a transverse endpoint field is (19). B(−κ000)((v,u),(0,z))=κv(z0)22.B_{\left(\begin{smallmatrix}-\kappa&0\\0&0\end{smallmatrix}\right)} \bigl((v,u),(0,z)\bigr)=\frac{\kappa v(z_0)^2}{2}.(19)✦

Fix 0<δ≤1/20<\delta\leq1/2 and an endpoint choice ε\varepsilon. Let UU be an open set of times containing the closed interval with endpoints a,ba,b. Suppose γ\gamma is a geodesic for gεg_\varepsilon on UU, the field JJ is smooth on UU, and 0≤γ0(s)≤10\leq\gamma_0(s)\leq1 throughout the closed interval. The intrinsic index satisfies (20). Igε(J;[a,b])=Eε,δ[a,b]−(Bε,δ(b)−Bε,δ(a)).I_{g_\varepsilon}(J;[a,b]) =E_{\varepsilon,\delta}[a,b] -\bigl(B_{\varepsilon,\delta}(b)-B_{\varepsilon,\delta}(a)\bigr).(20)✦

What Is Checked

The algebraic statements concern split vectors and real transverse matrices with the specified inverse and symmetry conditions. The interval statements concern continuous coordinate data satisfying (16), and smooth geodesics and fields in the two endpoint-framed neck metrics.

The triangular frame depends smoothly on smooth nonvanishing radii and smooth shear.✦ Pullback tangent inner products by a smooth map are smooth sections of the bilinear-form bundle.✦ A smooth field of bilinear forms on a vector space gives a smooth section of the bundle of bilinear forms on its tangent spaces.✦ For a smooth moving frame K(x)K(x), the bilinear form (u,w)↦⟨K(x)u,K(x)w⟩(u,w)\mapsto\langle K(x)u,K(x)w\rangle depends smoothly on xx.✦