Research map / Federated learning and privacy / Differential privacy
Differential Privacy without Sensitivity
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Use Lipschitz and convex loss with log-concave prior (Gibbs posterior analysis via LSI) instead of bounded loss sensitivity for (eps, delta)-DP | Gibbs posterior / exponential mechanism · theoretical analysis, unbounded parameter space | maximum admissible beta for (eps, delta)-DP | Helped |
| Larger prior precision m_pi (more shrinkage) in the Gibbs posterior | Gibbs posterior with Lipschitz convex loss · theoretical analysis (Corollary 8) | upper bound on beta for (eps, delta)-DP, proportional to sqrt(m_pi) | Helped |
| Increase dataset size n in strongly convex loss setting | Gibbs posterior with strongly convex loss (Theorem 10) · theoretical analysis | tail bound right-hand side in (9) | Helped |
| Approximate Gibbs sampling via Langevin Monte Carlo with total variation error gamma | Gibbs posterior sampler · theoretical analysis (Proposition 13) | privacy guarantee (eps, delta + (1+e^eps) gamma) | No clear effect |
Closest papers by meaning
- Benefits and Pitfalls of the Exponential Mechanism with Applications to Hilbert Spaces and Functional PCA (ICML 2019 · similarity 0.91)
- Tractable MCMC for Private Learning with Pure and Gaussian Differential Privacy (ICLR 2024 · similarity 0.89)
- Approximate Differential Privacy of the $\ell_2$ Mechanism (ICML 2025 · similarity 0.89)
- Beyond the Calibration Point: Mechanism Comparison in Differential Privacy (ICML 2024 · similarity 0.89)
- Privacy Guarantees in Posterior Sampling under Contamination (NeurIPS 2026 · similarity 0.88)
- On Inherent Privacy of Posterior Sampling: A Unified R\'enyi-Divergence Framework (NeurIPS 2026 · similarity 0.88)
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.