{
 "artifacts": null,
 "category": "architecture",
 "description": "Construct a Fourier layer whose active frequencies lie on several nonparallel polygonal patches or thin annular sectors, and cap repeated difference vectors generated by pairs of patches. The bounded-multiplicity geometry limits how many input frequency pairs can contribute to the same output frequency, potentially reducing spectral aliasing and gradient variance in nonlinear Fourier mixing.",
 "formulas_latex": [
  "$$|\\xi-\\zeta|\\geqslant c_{P}A-2$$",
  "$$\\#\\Bigl\\{(\\xi,\\zeta)\\in\\Gamma_{i,R}^{(A)}\\times\\Gamma_{j,R}^{(A)}:\\zeta-\\xi=k\\Bigr\\}\\lesssim_{P}1$$",
  "$$m(k):=\\#\\{(i,j,\\xi,\\zeta):\\xi\\in\\Gamma_i,\\zeta\\in\\Gamma_j,\\zeta-\\xi=k\\},\\qquad \\mathcal L_{\\mathrm{coll}}=\\sum_{k\\in\\mathcal K}[m(k)-M]_+^2$$",
  "$$\\widehat{(fg)}(k)=\\sum_{\\xi+\\zeta=k}\\widehat f(\\xi)\\widehat g(\\zeta),\\qquad \\widehat h(k)=\\sum_{\\zeta-\\xi=k}W_{\\xi,\\zeta}\\widehat f(\\xi)\\widehat g(\\zeta)$$"
 ],
 "id": 2756,
 "implementation": "(1) Integration point: modify the spectral mixing operation in a 2D Fourier neural operator, Fourier MLP, or spectral CNN. Represent each feature map by an H by W FFT, retain frequencies in a band near radius R, and divide the retained lattice frequencies into 3 to 8 patches associated with nonparallel sides of a fixed convex polygon such as a square or hexagon. Use separate complex weights for cross-patch products and then apply an inverse FFT followed by the usual residual block. (2) Pseudocode: precompute Gamma_i as the integer frequencies within radial distance h of R S_i and at least A tangential units from the side endpoints. For each pair xi in Gamma_i and zeta in Gamma_j with i not equal to j, compute k equal to zeta minus xi and insert the pair into a dictionary indexed by k. Either retain only the first M pairs per key or attach a learnable gate and optimize the collision penalty. During the forward pass, compute FFTs F and G, then set H[k] equal to the sum of W[xi,zeta] F[xi] G[zeta] over retained pairs with zeta minus xi equal to k. Add IFFT(H) to the residual stream. (3) Compute exactly from the geometry: patch membership, difference keys, multiplicities, and the separation test |xi-zeta| \u003e= c_P A - 2. Estimate empirically the useful cap M, the value of c_P for the chosen polygon, spectral energy retention, and conditioning. (4) First experiment: train a 6-layer Fourier neural operator on a 64 by 64 periodic Burgers or Navier-Stokes dataset. Compare dense spectral multiplication, random frequency partitions, and polygonal transverse patches at equal parameter count and FFT cost. Measure validation error, loss at equal FLOPs, gradient-norm variance, high-frequency energy, and divergent-run frequency. A successful result would be lower validation error or substantially fewer unstable runs at the same retained-mode budget.",
 "math_summary": "For a convex polygon P, let S_i and S_j be two nonparallel sides, and let Gamma_{i,R}^{(A)} and Gamma_{j,R}^{(A)} be lattice-frequency subsets lying near the corresponding A-trimmed neighborhoods of the dilated sides R S_i and R S_j. The paper proves the separation estimate |xi-zeta| \u003e= c_P A - 2 for every pair (xi,zeta) in these sets, where c_P \u003e 0 depends only on P and A is the tangential distance from the side endpoints. It also proves that, for every difference vector k in Z^2, the number of pairs satisfying zeta-xi=k is bounded by a constant depending only on P. For a neural spectral product, define m(k) as the number of retained frequency pairs producing difference k. The adaptation either enforces m(k) \u003c= M by hard pair selection or penalizes excess collisions with sum_k [m(k)-M]_+^2. The mathematical asset is bounded multiplicity: each output difference frequency receives only a controlled number of cross-patch interactions.",
 "math_tags": [
  "harmonic-analysis",
  "geometry",
  "combinatorics",
  "linear-algebra"
 ],
 "ml_areas": [
  "mlp",
  "cnn",
  "regularization",
  "training"
 ],
 "paper": {
  "arxiv_id": "2608.27544",
  "arxiv_url": "https://arxiv.org/abs/2608.27544",
  "summary_what_math_gives_to_ml": "The paper develops explicit geometric decompositions of thin Fourier annuli at the curvature scale R^{-1/2}, together with discrete bounds on overlaps and difference multiplicities. Its polygonal-boundary result shows that frequency pairs from nonparallel sides have uniformly bounded multiplicity for every fixed difference vector. This suggests a structured Fourier neural layer whose learnable frequency supports are partitioned into transverse sectors or polygonal patches, limiting many-to-one collisions in spectral convolutions. The most practical transfer is an anti-aliasing and variance-control mechanism for Fourier layers and neural operators, rather than a direct use of the PDE observability theorem.",
  "title": "Quantitative Uniqueness and Rough Damping on $\\mathbb T^2$",
  "year": "2026"
 },
 "ratings": {
  "difficulty": 6,
  "novelty": 7,
  "usefulness": 5
 },
 "solves": [
  "stability",
  "accuracy",
  "scalability"
 ],
 "title": "Transverse Fourier Collision Control",
 "url": "https://synthcore.org/idea/2756/transverse-fourier-collision-control",
 "verification": {
  "peer_reviewed": false,
  "status": "unverified",
  "status_label": "Unverified",
  "verdict_source": "deterministic test code (paired-seed permutation statistics)"
 }
}