Skip to the content

Math 2.0

Resources / preprint

On Inherent Privacy of Posterior Sampling: A Unified R\'enyi-Divergence Framework NM6Z2R61CN

URI	`m2://res/NM6Z2R61CN`
Kind	preprint
Creators	Ruyu Zhou, Fang Liu
Date	2026-10-14
Subjects	arxiv stat.ML, arxiv cs.LG, arxiv math.ST, arxiv stat.TH
Scope	border

Expressions

e1: version, preprint

Date	2026-10-14
Ids	arXiv 2610.19573

Items in this resource

The mentions of the inventory that name this resource. Each name links to the card of its inventory row.

Item	Kind	Where	Source
4ES6QBSQXJ	statement	lemma A.3	arxiv
Logistic regression with Gaussian prior	example	example 10	arxiv
RDP guarantees of OPS with sub-Gaussian log-likelihood ratio	statement	corollary 8	arxiv
Pure-DP guarantee	statement	corollary 6	arxiv
Relations between R\'enyi divergence and DP notions	statement	proposition 1	arxiv
Differential Privacy (DP	definition	definition 1	arxiv
GLMs with Gaussian prior	statement	proposition 9	arxiv
Exact characterization of R\'enyi divergence between neighboring posteriors	statement	theorem 3	arxiv
DS6XERS94C	statement	lemma A.2	arxiv
R\'enyi Differential Privacy (RDP	definition	definition 3	arxiv
RDP guarantees of OPS with sub-exponential log-likelihood ratio	statement	corollary 13	arxiv
J1AYRTP2EJ	statement	lemma	arxiv
KF96WEVE14	statement	lemma	arxiv
Gaussian-Gaussian conjugate pair with known variance	statement	proposition 11	arxiv
Marginal-Pointwise identity	statement	lemma 2	arxiv
Linear regression with Gaussian prior	statement	proposition 12	arxiv
Attainable RDP guarantees of OPS with uniformly bounded log-likelihood ratio	statement	corollary 5	arxiv
Privacy Loss Random Variable	definition	definition 2	arxiv
Categorical likelihood with arbitrary prior	example	example 7	arxiv

Catalogue records

arxiv: 2610.19573 (work)