AI Paper Atlas

Research map / Federated learning and privacy / Differential privacy

Unleashing the Power of Randomization in Auditing Differentially Private ML

Krishna Pillutla, Galen Andrew, Peter Kairouz, Hugh Brendan McMahan, Alina Oprea, Sewoong Oh

NeurIPS 2023 · Federated learning and privacy › Differential privacy · approach: Private online and user-level learning · 4 citations (OpenAlex)

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 changedWhereMeasuredEffect
audit with K=16 random canaries instead of a single canary (data poisoning)DP-SGD auditing · FMNIST/Linear, Purchase/MLP, eps 2-16sample complexity factor: 2.3x averageImproved
audit with K=16 random canaries instead of a single canary (random gradients)DP-SGD auditing · FMNIST/Linear, FMNIST/MLP, Purchase/MLPsample complexity factor: 3.0x averageImproved
LiDP auditing vs auditing DP across 12 settings per canary typeLiDP auditing · FMNIST, Purchase-100sample complexity factor: 11 out of the 12 settingsMixed
inject K canaries with 2nd-order Wilson interval vs single canaryGaussian mechanism auditing · Gaussian sum querytrials n to reach same eps lower bound: 4x to 16xImproved
reach eps lower bound 0.85LiDP + 2nd-order Wilson · Gaussian sum queryrequired trials: 65536 -> 4096Improved
reach eps lower bound 0.67LiDP + 2nd-order Wilson · Gaussian sum queryrequired trials: 4096 -> 1024Improved
increase number of canaries K (ablation)LiDP auditing with 2nd/4th-order CIs · Gaussian sum queryeps lower boundMixed
increase K (bias term) (ablation)LiDP auditing · Gaussian sum queryDelta BiasHurt
increase K (variance term) (ablation)LiDP auditing · Gaussian sum queryDelta VarImproved
use 4th-order instead of 2nd-order interval (ablation)LiDP auditing · Gaussian sum query, small epseps lower boundNo clear effect
increase dimension d (ablation)LiDP auditing · Gaussian sum queryeps lower boundImproved
4th-order vs 2nd-order Wilson, FMNIST/Linear data poisoning eps=2 (ablation)LiDP auditing · FMNIST/Linearsample complexity factor: 0.68 -> 0.42Hurt

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy