{
 "artifacts": null,
 "category": "optimization",
 "description": "Use a Goldfarb–Idnani-style active-set solver as a neural constrained layer or optimizer substep, but never trust a guessed active set solely because its linear system solved. Remove duplicate or dependent constraints, solve the reduced KKT system, and accept the result only after checking primal feasibility, dual sign conditions, and stationarity. This gives exact enforcement of linear inequalities and a diagnostic certificate when the constraint set is infeasible.",
 "formulas_latex": [
  "$$\\min_x\\;\\frac12x^\\top Gx-a^\\top x\\quad\\text{subject to}\\quad C^\\top x\\ge b,\\qquad G\\succ0.$$",
  "$$C_{\\mathcal A}^\\top G^{-1}C_{\\mathcal A}\\succ0\\quad\\Longleftrightarrow\\quad C_{\\mathcal A}\\text{ has full column rank}.$$",
  "$$Gx-a-C\\lambda=0,\\quad C^\\top x-b\\ge0,\\quad \\lambda\\ge0,\\quad \\lambda_i(C_i^\\top x-b_i)=0.$$",
  "$$Cy=0,\\quad y\\ge0,\\quad b^\\top y\u003e0\\quad\\Longrightarrow\\quad\\{x:C^\\top x\\ge b\\}=\\varnothing.$$"
 ],
 "id": 2910,
 "implementation": "Integrate this at the output of a neural module whose prediction must obey linear constraints: for example, a portfolio allocation head with $x\\ge0$ and $\\mathbf1^\\top x=1$, a control head with actuator bounds and rate inequalities, or a resource-allocation head with $Cx\\ge b$. At each forward pass, form a strictly convex QP with $G=\\operatorname{diag}(\\operatorname{softplus}(g)+\\epsilon)$ or a fixed SPD matrix, and let the network produce $a$. Initialize at the unconstrained minimizer $x_0=G^{-1}a$ and maintain an active set $\\mathcal A$. Use this loop: (1) find the most violated constraint $i=\\arg\\min_j(C_j^\\top x-b_j)$ and add it if its violation is below $-\\tau$; (2) deduplicate and rank-reveal $C_{\\mathcal A}$ using QR with threshold $\\rho$, discarding dependent constraints or retaining one representative from each dependent group; (3) solve the equality-constrained KKT system $$\\begin{bmatrix}G\u0026-C_{\\mathcal A}\\\\C_{\\mathcal A}^\\top\u00260\\end{bmatrix}\\begin{bmatrix}x\\\\\\lambda_{\\mathcal A}\\end{bmatrix}=\\begin{bmatrix}a\\\\b_{\\mathcal A}\\end{bmatrix};$$ (4) if any active multiplier is negative, remove the most negative one and repeat; (5) terminate only when the maximum violation is at most $\\tau$, the minimum multiplier is at least $-\\tau$, and the stationarity residual $\\|Gx-a-C\\lambda\\|_\\infty$ is at most $\\tau$. Backpropagate through the final linear solve using implicit differentiation; for an initial implementation, stop gradients through active-set decisions and differentiate only the KKT solve. Compute rank and residual thresholds empirically, while the KKT solution, feasibility checks, and multiplier checks come directly from the paper's mathematics. The first cheap experiment should use a small MLP on a synthetic portfolio dataset and compare this layer with softmax-plus-penalty and projected gradient descent at equal forward cost. Measure constraint violation, objective gap against a trusted CVXPY solution, gradient stability, wall-clock time, and training loss. A successful result is near-machine-precision feasibility, lower objective gap, and fewer failed or oscillatory training steps, especially after duplicating or nearly duplicating constraint columns.",
 "math_summary": "The paper studies the strictly convex quadratic program $$\\min_x\\;\\frac12x^\\top Gx-a^\\top x\\quad\\text{subject to}\\quad C^\\top x\\ge b,$$ where $G\\in\\mathbb R^{n\\times n}$ is symmetric positive definite, $a\\in\\mathbb R^n$, $C\\in\\mathbb R^{n\\times m}$, and $b\\in\\mathbb R^m$. For an active-set guess $\\mathcal A\\subseteq\\{1,\\ldots,m\\}$, the working-set matrix is $$C_{\\mathcal A}^\\top G^{-1}C_{\\mathcal A},$$ which is positive definite only if the selected columns of $C_{\\mathcal A}$ are linearly independent; arbitrary guessed active sets therefore do not inherit the safe principal-submatrix property of bound constraints. The KKT conditions are sufficient because $G\\succ0$: $$Gx-a-C\\lambda=0,\\qquad C^\\top x-b\\ge0,\\qquad \\lambda\\ge0,\\qquad \\lambda_i(C_i^\\top x-b_i)=0.$$ Here $x$ is the primal vector, $\\lambda$ are nonnegative multipliers, and $C_i$ is constraint column $i$. A candidate is certified by checking all four conditions, not by trusting the active-set guess. For infeasibility, a Farkas certificate is any $y\\ge0$ with $$Cy=0,\\qquad b^\\top y\u003e0,$$ because multiplying $C^\\top x\\ge b$ by $y^\\top$ would imply $0=y^\\top C^\\top x\\ge b^\\top y\u003e0$. The paper also gives the exact objective change along a step $x(t)=x_0+t z$: $$f(x(t))-f(x_0)=t\\left(\\frac t2+u_*\\right)z^\\top Gz,$$ where $f(x)=\\frac12x^\\top Gx-a^\\top x$, $z$ is the search direction, and $u_*$ is the directional linear coefficient after reducing the quadratic to one dimension; minimizing gives $t_*=-u_*$ when feasible.",
 "math_tags": [
  "optimization",
  "convex-analysis",
  "linear-algebra",
  "geometry"
 ],
 "ml_areas": [
  "optimizer",
  "loss",
  "regularization",
  "training"
 ],
 "paper": {
  "arxiv_id": "2608.30933",
  "arxiv_url": "https://arxiv.org/abs/2608.30933",
  "summary_what_math_gives_to_ml": "The paper exposes a useful distinction between guessing an active set and certifying a candidate solution for strictly convex quadratic programs. Its transferable asset is a rank-aware, certificate-producing solver: arbitrary inequality guesses are unsafe because the working-set matrix C_A^T G^{-1} C_A is positive definite only when the guessed constraint normals are linearly independent, but any candidate satisfying the KKT conditions is globally correct. This suggests replacing heuristic projection or penalty steps in constrained neural modules with small active-set QP solves that explicitly validate primal feasibility, dual signs, stationarity, and rank. The approach is most promising for differentiable constrained layers, portfolio or control heads, and neural optimizers enforcing linear resource or monotonicity constraints rather than as a generic replacement for Adam.",
  "title": "Goldfarb-Idnani Revisited:Invariants, Certificates, and the Limits of Guessing",
  "year": "2026"
 },
 "ratings": {
  "difficulty": 6,
  "novelty": 5,
  "usefulness": 5
 },
 "solves": [
  "stability",
  "accuracy",
  "sample-efficiency"
 ],
 "title": "Certified Rank-Aware QP Layer",
 "url": "https://synthcore.org/idea/2910/certified-rank-aware-qp-layer",
 "verification": {
  "peer_reviewed": false,
  "status": "unverified",
  "status_label": "Unverified",
  "verdict_source": "deterministic test code (paired-seed permutation statistics)"
 }
}