Research map / Federated learning and privacy / Differential privacy
Exponential-Wrapped Mechanisms: Differential Privacy on Hadamard Manifolds Made Practical
Yangdi Jiang, Xiaotian Chang, Lei Ding, Linglong Kong, Bei Jiang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| replace Riemannian Laplace with Exponential-Wrapped Gaussian mechanism | GDP Fréchet mean release · SPDM affine-invariant metric, m=30 | computation time (s): 6.85141 -> 0.00709 | Improved |
| replace Riemannian Laplace with Exponential-Wrapped Gaussian mechanism | GDP Fréchet mean release · SPDM affine-invariant metric, m=2 | computation time (s): 1.01774 -> 0.00248 | Improved |
| use EWG instead of RL under Log-Cholesky and Log-Euclidean metrics | GDP Fréchet mean release · SPDM Log-Cholesky/Log-Euclidean, d in {3,10,15} | Riemannian distance to Fréchet mean | Improved |
| use EWG instead of RL under affine-invariant metric, d=3 | GDP Fréchet mean release · SPDM affine-invariant, d=3, mu in [0.7, 2] | Riemannian distance to Fréchet mean | Hurt |
| use EWG instead of RL under affine-invariant metric, d=10,15 | GDP Fréchet mean release · SPDM affine-invariant, d in {10,15} | Riemannian distance to Fréchet mean | Improved |
| use EWG instead of RL on hyperbolic space | GDP Fréchet mean release · Hyperbolic space, d in {3,10,15} | Riemannian distance to Fréchet mean | Improved |
| use EWG instead of RL on OCTMNIST covariance descriptors | GDP Fréchet mean release · OCTMNIST classes 0-3, Log-Euclidean SPD5 | Riemannian distance to Fréchet mean | Improved |
Closest papers by meaning
- Differential Privacy Over Riemannian Manifolds (NeurIPS 2021 · similarity 0.91)
- Gaussian Differential Privacy on Riemannian Manifolds (NeurIPS 2023 · similarity 0.90)
- Shape And Structure Preserving Differential Privacy (NeurIPS 2022 · similarity 0.89)
- Approximate Differential Privacy of the $\ell_2$ Mechanism (ICML 2025 · similarity 0.87)
- 𝑓-Differential Privacy Filters: Validity and Approximate Solutions (NeurIPS 2026 · similarity 0.86)
- Differentially Private Set Representations (NeurIPS 2024 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
