Pith. sign in

REVIEW 6 cited by

Composition of Differential Privacy & Privacy Amplification by Subsampling

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2210.00597 v4 pith:CKFJU5IR submitted 2022-10-02 cs.CR cs.DScs.LG

Thomas Steinke This is my paper
classification cs.CRcs.DScs.LG
keywords privacydifferentialamplificationanalyseschaptercompositionprivatesubsampling
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

This chapter is meant to be part of the book "Differential Privacy for Artificial Intelligence Applications." We give an introduction to the most important property of differential privacy -- composition: running multiple independent analyses on the data of a set of people will still be differentially private as long as each of the analyses is private on its own -- as well as the related topic of privacy amplification by subsampling. This chapter introduces the basic concepts and gives proofs of the key results needed to apply these tools in practice.

Discussion (0). Sign in to comment.

Forward citations

Cited by 6 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Differentially Private Nonparametric Modal Learning with Applications to Regression and Clustering

    math.ST 2026-07 conditional novelty 7.0 of 10

    A private gradient-ascent algorithm estimates all density modes with nearly minimax-optimal error under differential privacy.

  2. Fixed-Composition Shuffle Asymptotics in the Full-Support Gaussian Regime

    cs.IT 2026-01 accept novelty 6.0 of 10

    For full-support shuffled binary-input local randomizers, neighboring shuffle histogram laws converge to a Gaussian shift with parameter sqrt(I_pi/n), giving JSD = I_pi/(8n) + O(n^{-2}) and explicit GDP privacy curves.

  3. Interpreting Differential Privacy in Terms of Disclosure Risk

    cs.CR 2025-07 accept novelty 6.0 of 10

    Shows that (epsilon,delta)-differential privacy bounds an adversary's posterior probability, posterior-to-prior ratio, and posterior-to-prior difference with high probability.

  4. Decentralized Optimization with Amplified Privacy via Efficient Communication

    eess.SY 2025-06 reject novelty 6.0 of 10

    Random activation and Top-k sparsification are claimed to amplify differential privacy in decentralized non-convex optimization, reducing required noise by a factor of the sparsification ratio times the square of the ...

  5. Federated Learning with Enhanced Privacy via Model Splitting and Random Client Participation

    cs.LG 2025-09 reject novelty 4.0 of 10

    MS-PAFL claims that adding DP noise only to a shared public submodel, combined with client and data subsampling, gives a central privacy loss of O(pqε) instead of O(ε).

  6. Large Language Model Powered Intelligent Urban Agents: Concepts, Capabilities, and Applications

    cs.MA 2025-07 conditional novelty 4.0 of 10

    The paper defines urban LLM agents, surveys their sensing, memory, reasoning, execution, and learning workflows, and organizes their applications across planning, transportation, environment, safety, and society.

Pith tools