AI Paper Atlas

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

Concept-Aware Privacy Mechanisms for Defending Embedding Inversion Attacks

Yu-Che Tsai, Hsiang Hsiao, Kuan-Yu Chen, Shou-De Lin

ICLR 2026 · Federated learning and privacy › Privacy attacks and secure learning · approach: Membership inference and auditing · 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
replace spherical Laplace noise with sensitivity-guided Mahalanobis noiseLapMech · STS12 eps=10leakage (%): 22.34 -> 19.31Improved
same, downstream utilityLapMech · STS12 eps=10downstream (%): 60.72 -> 65.27Improved
same, at large budget eps=40LapMech · STS12downstream (%): 73.98 -> 74.15No clear effect
use SPARSE with Sentence-T5 embeddingsLapMech · STS12 eps=10leakage (%): 31.71 -> 22.38Improved
use SPARSE against shallow MLC attackerLapMech · STS12 eps=5leakage (%): 19.39 -> 17.63Improved
SPARSE vs LapMech on STS14 at eps=30LapMech · STS14leakage (%): 11.22 -> 14.70Hurt
SPARSE vs LapMech on NFCorpus at eps=5LapMech · NFCorpusdownstream (%): 23.70 -> 19.94Hurt
increase mask regularization (lambda 1e-3 to 1e-2) (ablation)SPARSE · STS12 eps=10leakage (%): 19.31 -> 0.91Improved
increase mask regularization (lambda 1e-3 to 1e-2) (ablation)SPARSE · STS12 eps=10downstream (%): 65.27 -> 36.87Hurt
use white-box attribution instead of learned mask (ablation)SPARSE · STS12 eps=10leakage (%): 19.31 -> 12.01Improved
use SPARSE instead of PII removalRemovePII · STS12downstream (%): 59.47 -> 73.25Improved

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