{
 "artifacts": null,
 "category": "optimization",
 "description": "Replace ordinary parameter updates for a constrained neural network with an annealed reciprocal-manifold flow. Each differentiable inequality constraint remains strictly satisfied during the optimization trajectory, avoiding projection or a per-step quadratic program. This is most useful for safety-critical policy learning, bounded network outputs, parameter-budget constraints, or training with explicit robustness inequalities.",
 "formulas_latex": [
  "$$\\lambda(x)=-\\frac{k}{g(x)}$$",
  "$$F_k(x)=f(x)-k\\log\\bigl(-g(x)\\bigr),\\qquad g(x)\u003c0$$",
  "$$\\dot{x}=-\\nabla f(x)-\\sum_{i=1}^{m}\\lambda_i\\nabla g_i(x),\\qquad \\lambda_i=-\\frac{k}{g_i(x)}$$",
  "$$\\dot{\\lambda}_i=-\\frac{\\dot{k}}{g_i(x)}+\\frac{k}{g_i(x)^2}\\nabla g_i(x)^\\top\\dot{x}$$"
 ],
 "id": 2840,
 "implementation": "1. Integration point: modify the optimizer update applied to neural-network parameters \\(\\theta\\), or to policy parameters in constrained RL. Define a batch of differentiable inequalities \\(g_i(\\theta)\u003c0\\); examples include \\(\\|\\theta\\|_2^2-B\u003c0\\), an estimated worst-case loss minus a safety threshold, action-limit violations, or a differentiable FLOP/memory budget. Start from a strictly feasible initialization. 2. Pseudocode: set \\(k=k_0\\), compute the task loss \\(L(\\theta)\\), constraints \\(g_i\\), and gradients; set \\(\\lambda_i=-k/(g_i+\\epsilon)\\); form \\(v=-\\nabla_\\theta L-\\sum_i\\lambda_i\\nabla_\\theta g_i\\); choose an adaptive step \\(h\\), update \\(\\theta\\leftarrow\\theta+h v\\), then recompute every constraint. Use \\(k\\leftarrow\\max(k_{\\min},k\\exp(-h/\\tau))\\). Before accepting a step, cap \\(h\\) using \\(h\\le \\eta\\min_i[-g_i/(|\\nabla g_i^\\top v|+\\epsilon)]\\), with \\(0\u003c\\eta\u003c1\\); reject and halve \\(h\\) if any constraint is nonnegative. 3. The reciprocal multiplier, barrier force, and annealing schedule come from the paper; the step cap and finite-step rejection are engineering safeguards because the guarantee concerns continuous time. Estimate constraint gradients with automatic differentiation and optionally clip \\(\\lambda_i\\) to prevent floating-point overflow. 4. First experiment: train a small PPO policy on a constrained continuous-control task such as Safety Gymnasium, comparing Adam, projected-gradient updates, and this optimizer at equal environment interactions. Log task return, cumulative constraint violation, minimum constraint margin, rejected steps, and wall-clock cost. Success means zero or substantially lower intermediate violations with comparable final return, and fewer expensive safety QP calls than the projected baseline. A second cheap test can impose a differentiable parameter-norm budget during CIFAR-10 MLP training and measure accuracy at fixed budget and the fraction of steps crossing the boundary.",
 "math_summary": "The paper defines the reciprocal multiplier manifold for an inequality constraint \\(g(x)\u003c0\\) as \\(\\lambda(x)=-k/g(x)\\), where \\(x\\) is the optimization state, \\(g:\\mathbb{R}^d\\to\\mathbb{R}\\) is a differentiable constraint function, \\(k\u003e0\\) is the annealed barrier parameter, and \\(\\lambda\u003e0\\) is its associated multiplier. The equivalent logarithmic-barrier objective is \\(F_k(x)=f(x)-k\\log(-g(x))\\), where \\(f\\) is the neural-network loss; its gradient flow is \\(\\dot{x}=-\\nabla f(x)-\\lambda\\nabla g(x)\\). For multiple constraints \\(g_i(x)\u003c0\\), use \\(\\lambda_i=-k/g_i(x)\\) and \\(\\dot{x}=-\\nabla f(x)-\\sum_i\\lambda_i\\nabla g_i(x)\\). The exact tangent multiplier dynamics induced by the manifold are \\(\\dot{\\lambda}_i=-\\dot{k}/g_i+(k/g_i^2)\\nabla g_i(x)^\\top\\dot{x}\\). The reciprocal relation supplies the repulsive force that diverges near a boundary; annealing \\(k\\to0\\) reduces barrier bias and targets the original constrained optimum. In code, the algebraic relation \\(\\lambda_i=-k/g_i\\) is safer than independently integrating \\(\\lambda_i\\), while the differentiated update is useful for testing the invariant-manifold dynamics.",
 "math_tags": [
  "optimization",
  "dynamical-systems",
  "control-theory",
  "convex-analysis"
 ],
 "ml_areas": [
  "optimizer",
  "training",
  "rl",
  "regularization"
 ],
 "paper": {
  "arxiv_id": "2608.29628",
  "arxiv_url": "https://arxiv.org/abs/2608.29628",
  "summary_what_math_gives_to_ml": "The paper's transferable asset is a reciprocal primal-dual representation in which an inequality constraint g(x)\u003c0 is paired with a positive multiplier through the explicit manifold \\(\\lambda=-k/g(x)\\). This converts constrained optimization into a barrier-like dynamical system whose state can remain feasible without projection or a per-step quadratic-program safety layer, while annealing \\(k\\to0\\) removes the barrier bias. A promising neural-network use is a feasibility-preserving optimizer for differentiable safety, norm, resource, or robustness constraints on network parameters or outputs. The main engineering risk is that the guarantee is continuous-time; discrete integration requires adaptive step sizes or a division-free coordinate implementation to avoid crossing the constraint boundary.",
  "title": "Reciprocal-Manifold Annealed KKT Flows for Constrained Optimization: Application to the Nonconvex AC Optimal Power Flow",
  "year": "2026"
 },
 "ratings": {
  "difficulty": 5,
  "novelty": 5,
  "usefulness": 6
 },
 "solves": [
  "stability",
  "accuracy",
  "generalization"
 ],
 "title": "Reciprocal Feasibility-Preserving Optimizer",
 "url": "https://synthcore.org/idea/2840/reciprocal-feasibility-preserving-optimizer",
 "verification": {
  "peer_reviewed": false,
  "status": "unverified",
  "status_label": "Unverified",
  "verdict_source": "deterministic test code (paired-seed permutation statistics)"
 }
}