{
 "artifacts": null,
 "category": "regularization",
 "description": "For a coordinate-based neural network u_theta(x) solving a fully nonlinear second-order PDE, replace the raw quadratic-Hessian residual with the concave, homogeneous operator G(D_x^2 u_theta)=sqrt(sigma_2(D_x^2 u_theta)). Add differentiable barriers that keep the predicted Hessian inside the positive branch Gamma_2, preventing optimization from entering regions where the PDE operator is non-elliptic.",
 "formulas_latex": [
  "$$\\Gamma_{2}=\\{A\\in\\operatorname{Sym}(n):\\sigma_{1}(A)\u003e0,\\ \\sigma_{2}(A)\u003e0\\},$$",
  "$$G(A)=\\sqrt{\\sigma_{2}(A)},\\qquad \\sigma_2(A)=\\frac12\\left[(\\operatorname{tr}A)^2-\\operatorname{tr}(A^2)\\right],\\qquad G(D^2u)=g:=\\sqrt f,$$",
  "$$r_\\theta(x)=\\sqrt{\\max\\left\\{\\frac12\\left[(\\operatorname{tr}H_\\theta)^2-\\operatorname{tr}(H_\\theta^2)\\right],\\varepsilon\\right\\}}-g(x),\\qquad H_\\theta=D_x^2u_\\theta,$$",
  "$$L(\\theta)=\\mathbb E_{x\\sim\\rho}\\left[r_\\theta(x)^2+\\lambda_1\\frac{\\operatorname{softplus}(\\tau(\\delta-s_1(x)))}{\\tau}+\\lambda_2\\frac{\\operatorname{softplus}(\\tau(\\delta-s_2(x)))}{\\tau}\\right]+\\lambda_{bc}L_{bc},\\quad s_1=\\operatorname{tr}H_\\theta,\\ s_2=\\tfrac12(s_1^2-\\|H_\\theta\\|_F^2).$$"
 ],
 "id": 2831,
 "implementation": "(1) Integrate this into a coordinate MLP u_theta:R^n-\u003eR used as a PINN or implicit function model. At each interior collocation batch x_b, use automatic differentiation to compute the full input Hessian H_b=D_x^2u_theta(x_b); these are derivatives with respect to coordinates, not parameter Hessians. Compute s_1b=trace(H_b) and s_2b=0.5*(s_1b^2-sum_ij H_bij^2). Replace a raw residual sigma_2(H_b)-f(x_b) with r_b=sqrt(clamp(s_2b,min=epsilon))-sqrt(f(x_b)). Add the two cone barriers and the boundary-condition loss. (2) Pseudocode: `u=net(x); H=hessian(u,x); s1=trace(H); s2=0.5*(s1*s1-(H*H).sum()); r=sqrt(clamp(s2,min=eps))-sqrt(f(x)); b1=softplus(tau*(delta-s1))/tau; b2=softplus(tau*(delta-s2))/tau; loss=mean(r*r+lam1*b1+lam2*b2)+lam_bc*boundary_loss; optimizer.step(loss)`. Start in float64 with epsilon=1e-8 and delta=1e-4, then anneal lambda_1 and lambda_2 from zero to their target values during the first 10-20% of training. (3) Compute sigma_2, G, and cone membership exactly from each batch Hessian. Estimate empirical cone occupancy as the fraction of points satisfying s_1\u003edelta and s_2\u003edelta; log the 95th percentile of ||H_b||_F and residual spatial variation. (4) First test a two-dimensional Dirichlet problem on a disk or square using a manufactured smooth solution u*. Set f=sigma_2(D^2u*)+c with c chosen so f is strictly positive. Compare the proposed loss against the raw sigma_2(H)-f loss using the same four-layer, 128-width tanh MLP, collocation points, optimizer, and number of Hessian evaluations across five seeds. Measure PDE residual, boundary error, relative C^2 error, cone-violation rate, Hessian norm, loss smoothness, and wall-clock time. Success is lower cone-violation rate and interior C^2 error, with faster residual reduction at equal Hessian-evaluation budget; failure is no improvement or excessive autodiff overhead.",
 "math_summary": "For a symmetric matrix A in Sym(n) with eigenvalues lambda_1,...,lambda_n, sigma_1(A)=sum_i lambda_i=tr(A) and sigma_2(A)=sum_{i\u003cj} lambda_i lambda_j=1/2[(tr A)^2-tr(A^2)]. The admissible cone is Gamma_2={A:sigma_1(A)\u003e0 and sigma_2(A)\u003e0}. The paper states that G(A)=sqrt(sigma_2(A)) is elliptic, concave, and homogeneous of degree one on Gamma_2, meaning G(tA)=tG(A) for t\u003e0. The target equation sigma_2(D^2u)=f\u003e0 is equivalently G(D^2u)=g=sqrt(f). For a neural network, H_theta(x)=D_x^2u_theta(x), s_1(x)=tr(H_theta(x)), and s_2(x)=1/2[(tr H_theta(x))^2-tr(H_theta(x)^2)]. Use residual r_theta(x)=sqrt(max(s_2(x),epsilon))-g(x). Enforce the cone with b_1=softplus(tau(delta-s_1))/tau and b_2=softplus(tau(delta-s_2))/tau, where epsilon\u003e0 prevents invalid square roots, delta\u003e0 is a safety margin, and tau controls barrier sharpness. The training objective is L=E[r_theta^2+lambda_1 b_1+lambda_2 b_2]+lambda_bc L_bc. The transferable assets are the elliptic branch restriction, concavity of G, and degree-one scaling.",
 "math_tags": [
  "pde",
  "convex-analysis",
  "spectral-theory",
  "geometry"
 ],
 "ml_areas": [
  "mlp",
  "loss",
  "regularization"
 ],
 "paper": {
  "arxiv_id": "2608.29484",
  "arxiv_url": "https://arxiv.org/abs/2608.29484",
  "summary_what_math_gives_to_ml": "The paper identifies the positive branch of the quadratic Hessian operator as a useful nonlinear spectral map: G(A)=sqrt(sigma_2(A)) is elliptic, concave, and positively homogeneous on the cone Gamma_2. These properties can be transferred to coordinate-based neural networks that solve fully nonlinear second-order PDEs, where unconstrained Hessian predictions can leave the elliptic branch and destabilize training. A practical adaptation is to use G(D_x^2 u_theta) as the PDE residual and add a differentiable barrier enforcing the Gamma_2 conditions. The regularity theorem motivates monitoring interior Hessian magnitude and variation, although it does not itself provide a general neural-network training guarantee.",
  "title": "Interior $C^{2,α}$ Regularity for the Quadratic Hessian Equation",
  "year": "2026"
 },
 "ratings": {
  "difficulty": 5,
  "novelty": 7,
  "usefulness": 5
 },
 "solves": [
  "stability",
  "accuracy",
  "generalization"
 ],
 "title": "Quadratic-Hessian cone regularizer",
 "url": "https://synthcore.org/idea/2831/quadratic-hessian-cone-regularizer",
 "verification": {
  "peer_reviewed": false,
  "status": "unverified",
  "status_label": "Unverified",
  "verdict_source": "deterministic test code (paired-seed permutation statistics)"
 }
}