Multi-distribution Rényi divergences are positive integrals of coincidence divergences C_α over four strata (simplex interior, mixed-sign cones, tropical boundary, KL edges).
Rényi divergence and Kullback–Leibler divergence
5 Pith papers cite this work, alongside 1,529 external citations. Polarity classification is still indexing.
5
Pith papers citing it
1,529
external citations · external index
representative citing papers
ALU uses public data to suppress unlearning cost quadratically while characterizing distribution mismatch effects, enabling mass unlearning with maintained utility.
Colored top-m-to-random shuffles on wreath products have explicit total-variation, separation, and L^q cutoff profiles driven by a Poisson limit for never-chosen labels.
A tight context-aware leakage bound for general linear queries is derived under a prior probability lower bound, proven strictly tighter than differential privacy and converging to it as the bound approaches zero.
Derives f-divergence change-of-measure inequalities via Legendre transform to extend PAC-Bayes bounds beyond the classical Donsker-Varadhan setting.
citing papers explorer
-
All you need is log
Multi-distribution Rényi divergences are positive integrals of coincidence divergences C_α over four strata (simplex interior, mixed-sign cones, tropical boundary, KL edges).
-
Unlearning with Asymmetric Sources: Improved Unlearning-Utility Trade-off with Public Data
ALU uses public data to suppress unlearning cost quadratically while characterizing distribution mismatch effects, enabling mass unlearning with maintained utility.
-
Explicit cutoff profiles for colored top-$m$-to-random shuffles
Colored top-m-to-random shuffles on wreath products have explicit total-variation, separation, and L^q cutoff profiles driven by a Poisson limit for never-chosen labels.
-
Context-aware Privacy Bounds for Linear Queries
A tight context-aware leakage bound for general linear queries is derived under a prior probability lower bound, proven strictly tighter than differential privacy and converging to it as the bound approaches zero.
-
Change of measure through the Legendre transform
Derives f-divergence change-of-measure inequalities via Legendre transform to extend PAC-Bayes bounds beyond the classical Donsker-Varadhan setting.