{
 "artifacts": null,
 "category": "geometry",
 "description": "Replace Euclidean or unprojected covariance averaging with a projected Bures-Wasserstein barycenter layer. Each unit-step barycenter update is followed by eigenvalue clipping into \\([\\alpha,\\beta]\\), preserving positive definiteness and preventing ill-conditioning without an additional eigendecomposition.",
 "formulas_latex": [
  "$$\\mathrm{B}(P,Q)=\\mathrm{d}^{2}_{\\mathrm{BW}}(P,Q):=\\operatorname{Tr}[P+Q]-2\\mathrm{F}(P,Q),\\qquad \\mathrm{F}(P,Q):=\\operatorname{Tr}\\!\\left[\\sqrt{P^{1/2}QP^{1/2}}\\right]=\\left\\|P^{1/2}Q^{1/2}\\right\\|_{1}.$$",
  "$$\\Phi(S)=\\frac12\\sum_{i=1}^{m}w_i\\,\\mathrm{B}(S,A_i),\\qquad w_i\\ge0,\\quad \\sum_{i=1}^{m}w_i=1.$$",
  "$$\\Pi_{[\\alpha,\\beta]}(X)=U\\operatorname{diag}\\!\\big(\\operatorname{clip}(\\lambda_1,\\alpha,\\beta),\\ldots,\\operatorname{clip}(\\lambda_d,\\alpha,\\beta)\\big)U^{\\top},\\quad X=U\\operatorname{diag}(\\lambda_1,\\ldots,\\lambda_d)U^{\\top}.$$",
  "$$S_{t+1}=\\Pi_{[\\alpha,\\beta]}(T(S_t)),\\qquad d_{\\mathrm{BW}}(S_t,S_\\star)\\lesssim(1-\\kappa^{-3/2})^t d_{\\mathrm{BW}}(S_0,S_\\star).$$"
 ],
 "id": 3150,
 "implementation": "(1) Integration point: add this as a differentiable covariance-aggregation module in a CNN, graph neural network, or set encoder. For each sample, form SPD feature covariances \\(A_i=H_i^\\top H_i/n+\\epsilon I\\), where \\(H_i\\) is a patch, token-neighborhood, or augmentation feature matrix. Replace arithmetic pooling with \\(K\\) projected BW barycenter iterations. (2) Pseudocode: initialize \\(S\\leftarrow\\sum_iw_iA_i\\). For \\(t=1,\\ldots,K\\), compute \\(R_i=(S^{1/2}A_iS^{1/2})^{1/2}\\), form the unit-step proposal \\(X\\leftarrow(\\sum_iw_iR_i)^2\\), eigendecompose \\(X=U\\operatorname{diag}(\\lambda)U^\\top\\), and update \\(S\\leftarrow U\\operatorname{diag}(\\operatorname{clip}(\\lambda,\\alpha,\\beta))U^\\top\\). Feed \\(S\\), or \\(\\log S\\), to the next layer and backpropagate through the matrix operations with stabilized eigendecomposition gradients. (3) Compute the BW update and clipping exactly; choose \\(\\alpha\\) and \\(\\beta\\) from running lower and upper eigenvalue quantiles of the inputs, with gradients stopped through these bounds. Estimate the empirical contraction by measuring successive \\(d_{\\mathrm{BW}}(S_t,S_{t-1})\\); the theorem supplies the non-expansiveness and claimed rate, while task improvement is empirical. (4) First experiment: compare arithmetic pooling, log-Euclidean pooling, unprojected BW iterations, and projected BW iterations on CIFAR-100 using a small ResNet and covariance features, with \\(K\\in\\{3,5,10\\}\\). Measure validation accuracy, barycenter objective, condition numbers, NaN or gradient-failure frequency, and loss versus eigendecomposition FLOPs. Success is lower instability and faster objective decrease at equal compute, with no extra eigendecomposition relative to the unprojected BW baseline.",
 "math_summary": "For positive semidefinite matrices \\(P,Q\\in\\mathrm{PSD}(d)\\), the squared Bures-Wasserstein distance is \\(\\mathrm{B}(P,Q)=\\operatorname{Tr}(P+Q)-2\\mathrm{F}(P,Q)\\), where \\(\\mathrm{F}(P,Q)=\\operatorname{Tr}[(P^{1/2}QP^{1/2})^{1/2}]=\\|P^{1/2}Q^{1/2}\\|_1\\). Given SPD inputs \\(A_1,\\ldots,A_m\\) and weights \\(w_i\\ge0\\) with \\(\\sum_iw_i=1\\), the BW barycenter minimizes \\(\\Phi(S)=\\frac12\\sum_iw_i\\mathrm{B}(S,A_i)\\). The paper's projection lemma states that for \\(X=U\\operatorname{diag}(\\lambda_1,\\ldots,\\lambda_d)U^\\top\\), \\(\\Pi_{[\\alpha,\\beta]}(X)=U\\operatorname{diag}(\\operatorname{clip}(\\lambda_j,\\alpha,\\beta))U^\\top\\) is the closed-form BW projection onto \\(\\mathcal C_{\\alpha,\\beta}=\\{S:\\alpha I\\preceq S\\preceq\\beta I\\}\\), and it is non-expansive: \\(d_{\\mathrm{BW}}(\\Pi(X),\\Pi(Y))\\le d_{\\mathrm{BW}}(X,Y)\\). Applying this projection to the unit-step BW fixed-point map \\(T\\), \\(S_{t+1}=\\Pi_{[\\alpha,\\beta]}(T(S_t))\\), gives the claimed rate \\((1-\\kappa^{-3/2})\\) when the ensemble condition number is \\(\\kappa\\) and inputs lie in the spectral interval.",
 "math_tags": [
  "geometry",
  "optimal-transport",
  "linear-algebra",
  "spectral-theory"
 ],
 "ml_areas": [
  "embedding",
  "cnn",
  "graph-nn",
  "optimizer"
 ],
 "paper": {
  "arxiv_id": "2609.03762",
  "arxiv_url": "https://arxiv.org/abs/2609.03762",
  "summary_what_math_gives_to_ml": "The paper gives a concrete stability mechanism for iterative optimization over symmetric positive-definite matrices: after each unit-step Bures-Wasserstein update, clip eigenvalues into a fixed interval. The transferable asset is that spectral clipping is the exact closed-form projection in the Bures-Wasserstein metric and is non-expansive, while reusing an eigendecomposition already required by the update. This suggests projected Bures covariance-pooling layers and projection-based optimizers for learned SPD parameters. The claimed convergence rate is dimension-independent and has iteration complexity scaling as \\(O(\\kappa^{3/2}\\log(1/\\varepsilon))\\).",
  "title": "Projected Riemannian Gradient Descent for the Bures-Wasserstein Barycenter: Dimension-Independent Linear Convergence at Unit Step Size",
  "year": "2026"
 },
 "ratings": {
  "difficulty": 5,
  "novelty": 6,
  "usefulness": 7
 },
 "solves": [
  "stability",
  "accuracy",
  "speedup"
 ],
 "title": "Projected Bures Covariance Pooling",
 "url": "https://synthcore.org/idea/3150/projected-bures-covariance-pooling",
 "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
   }
  }
 }
}