{
 "artifacts": null,
 "category": "architecture",
 "description": "Construct a graph or sequence feature layer whose output satisfies a learned or fixed finite-support convolution equation, rather than allowing arbitrary features. For grid data, this is a hard local harmonicity or stencil constraint; for irregular graphs, it is imposed with a sparse incidence operator. The constraint can be exact through nullspace projection or soft through a residual penalty, and should improve robustness when the target signal is locally smooth or obeys known conservation relations.",
 "formulas_latex": [
  "$$(f*x_n)(g)=\\sum_{h\\in G}f(gh^{-1})x_n(h)=0,$$",
  "$$T(x):=\\sum_{n=1}^{\\infty}x(n)f_n,$$",
  "$$|\\nabla T(x)(e)|\\leq\\sum_{n=1}^{\\infty}|x(n)\\nabla f_n(e)|\\leq\\|x\\|_{\\ell_{\\infty}}\\sum_{n=0}^{\\infty}2^{-2n}\u003c\\infty,$$",
  "$$H_{\\mathrm{con}}=PH,\\qquad P=I-C^{\\dagger}C,\\qquad CH_{\\mathrm{con}}=0.$$"
 ],
 "id": 126,
 "implementation": "(1) Integration point: insert the module after a graph-convolution or Transformer token-mixing block and before the next nonlinear MLP. Let $H\\in\\mathbb{R}^{N\\times D}$ be node or token features. Build a sparse local constraint matrix $C$ from a fixed stencil, such as the graph Laplacian $C=L$, a discrete biharmonic stencil, or a user-specified signed neighborhood kernel. For a grid, implement $CH$ as a depthwise convolution; for a graph, use sparse matrix multiplication. (2) Pseudocode: `H = mixer(X)`; `R = C @ H`; either use `Hc = H - lambda * C.T @ solve(C @ C.T + eps*I, R)` as an approximate projection, or precompute a low-rank/nullspace basis $U$ and use `Hc = U @ (U.T @ H)`; then `Y = mlp(Hc)`. Add `loss += alpha * mean((C @ Hc)**2)` if using a soft constraint. Estimate the edge-gradient statistic with `G_e = Hc[src]-Hc[dst]` and optionally add `beta * mean(norm(G_e,2))`. (3) Mathematics computed exactly: the local convolution residual $CH$, the projection formula, and the gradient-bound surrogate. Quantities such as the pseudoinverse, rank, and suitable $\\alpha,\\beta$ are estimated numerically; use conjugate gradients rather than forming $C^{\\dagger}$ for large graphs. (4) First experiment: train a 4-layer GraphSAGE or GCN on Cora, Citeseer, and a noisy-grid image-classification task, comparing an unconstrained model, soft residual regularization, and projected features at equal parameter count. Measure validation accuracy under feature noise, the mean constraint residual $\\|CH\\|_F/\\|H\\|_F$, and gradient explosion during training. The idea succeeds if projection improves corruption accuracy or calibration without reducing clean accuracy by more than 1%, while keeping the residual near numerical zero and adding less than 20% wall-clock cost.",
 "math_summary": "The paper describes translation-invariant subspaces on a group G through convolution equations of the form $(f*x_n)(g)=\\sum_{h\\in G}f(gh^{-1})x_n(h)=0$, where $f:G\\to\\mathbb{R}$ is a Lipschitz function, $x_n:G\\to\\mathbb{R}$ is a finite or summable constraint kernel, $g\\in G$ is the translated location, and $*$ is group convolution. The associated constrained space is the set of Lipschitz functions annihilated by every translated kernel. The construction in the paper defines $T(x)=\\sum_{n=1}^{\\infty}x(n)f_n$ and controls its edge gradient using $|\\nabla T(x)(e)|\\leq\\|x\\|_{\\ell_\\infty}\\sum_{n=0}^{\\infty}2^{-2n}$, where $e$ is a graph edge and $\\nabla f(e)$ is the endpoint difference across that edge. We adapt the first result to a finite stencil matrix $C$ and the second principle to a feature-Lipschitz regularizer: constrain $CH=0$ and penalize or normalize the edge-gradient norm. For a batch of node features $H\\in\\mathbb{R}^{N\\times D}$, $C\\in\\mathbb{R}^{N\\times N}$ is a sparse matrix assembled from translated local kernels; on a $d$-dimensional grid, $C$ is a convolution operator. The exact nullspace projection is $P=I-C^{\\dagger}C$, where $C^{\\dagger}$ is the Moore-Penrose pseudoinverse, so $H_{\\mathrm{con}}=PH$ satisfies $CH_{\\mathrm{con}}=0$ up to numerical error.",
 "math_tags": [
  "harmonic-analysis",
  "functional-analysis",
  "metric-geometry",
  "linear-algebra"
 ],
 "ml_areas": [
  "graph-nn",
  "cnn",
  "regularization"
 ],
 "paper": {
  "arxiv_id": "2608.24143",
  "arxiv_url": "https://arxiv.org/abs/2608.24143",
  "summary_what_math_gives_to_ml": "The paper characterizes translation-invariant, pointwise-closed subspaces of graph Lipschitz functions through finite local constraints, which become convolution equations on groups. The transferable asset is a way to impose exact local harmonic or annihilation constraints on learned node or token functions while retaining Lipschitz-gradient control. Its strongest practical use is not the Banach-space dichotomy itself, but a constrained graph or sequence feature module whose outputs lie in the nullspace of selected finite convolution stencils; this can provide a controllable inductive bias for denoising, smoothing, or equivariant representation learning.",
  "title": "Invariant pointwise closed subspaces of Lipschitz spaces and their preduals",
  "year": "2026"
 },
 "ratings": {
  "difficulty": 6,
  "novelty": 6,
  "usefulness": 5
 },
 "solves": [
  "accuracy",
  "stability",
  "generalization"
 ],
 "title": "Convolution-Nullspace Feature Layer",
 "url": "https://synthcore.org/idea/126/convolution-nullspace-feature-layer",
 "verification": {
  "peer_reviewed": false,
  "status": "unverified",
  "status_label": "Unverified",
  "verdict_source": "deterministic test code (paired-seed permutation statistics)",
  "verification_axes": {
   "benchmark_mechanism": {
    "confirmed": null,
    "tested": false
   },
   "practical_benchmark": {
    "beats_baseline": null,
    "tested": false
   },
   "toy_mechanism_gate": {
    "confirmed": null,
    "tested": false
   }
  }
 }
}