AI Paper Atlas

Research map / Federated learning and privacy / Privacy attacks and secure learning

Posthoc privacy guarantees for collaborative inference with modified Propose-Test-Release

Abhishek Singh, Praneeth Vepakomma, Vivek Sharma, Ramesh Raskar

NeurIPS 2023 · Federated learning and privacy › Privacy attacks and secure learning · approach: Encryption-based private learning · 2 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
add contrastive and noisy regularization to ARL (ablation)ARL obfuscator · MNIST eps=1test accuracy: 0.329 -> 0.514Improved
add contrastive and noisy regularization to ARL (ablation)ARL obfuscator · FMNIST eps=1test accuracy: 0.214 -> 0.371Improved
add contrastive and noisy regularization to ARL (ablation)ARL obfuscator · UTKFace eps=1test accuracy: 0.605 -> 0.650Improved
add noisy regularization to ARL only (ARL-N, no contrastive) (ablation)ARL obfuscator · MNIST eps=10test accuracy: 0.882 -> 0.325Hurt
add adversarial representation learning to C-N (ablation)C-N obfuscator · FMNIST eps=10test accuracy: 0.599 -> 0.695Improved
add adversarial representation learning to C-N (ablation)C-N obfuscator · MNIST eps=10test accuracy: 0.913 -> 0.912No clear effect
add adversarial representation learning to C-N (ablation)C-N obfuscator · UTKFace eps=10test accuracy: 0.722 -> 0.700Hurt
replace ARL-C-N with plain Encoder at high privacy budgetObfuscator · MNIST eps=10test accuracy: 0.921 -> 0.912No clear effect
replace Encoder with ARL-C-N at strict privacyObfuscator · MNIST eps=1test accuracy: 0.428 -> 0.514Improved
add noisy regularization alone to plain pipeline vs ARL (ablation)Obfuscator · FMNIST eps=5test accuracy: 0.557 -> 0.310Hurt
increase adversarial weighting alpha from 0 to 0.9999 (ablation)Obfuscator · Local Lipschitz constantlocal Lipschitz constantImproved
use larger privacy budget eps (1 -> 10)ARL-C-N · MNISTtest accuracy: 0.514 -> 0.912Improved

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