English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

When learning from sensitive data, care must be taken to ensure that trainingalgorithms address privacy concerns. The canonical Private Aggregation ofTeacher Ensembles, or PATE, computes output labels by aggregating thepredictions of a (possibly distributed) collection of teacher models via avoting mechanism. The mechanism adds noise to attain a differential privacyguarantee with respect to the teachers' training data. In this work, we observethat this use of noise, which makes PATE predictions stochastic, enables newforms of leakage of sensitive information. For a given input, our adversaryexploits this stochasticity to extract high-fidelity histograms of the votessubmitted by the underlying teachers. From these histograms, the adversary canlearn sensitive attributes of the input such as race, gender, or age. Althoughthis attack does not directly violate the differential privacy guarantee, itclearly violates privacy norms and expectations, and would not be possible atall without the noise inserted to obtain differential privacy. In fact,counter-intuitively, the attack becomes easier as we add more noise to providestronger differential privacy. We hope this encourages future work to considerprivacy holistically rather than treat differential privacy as a panacea.

In Differential Privacy, There is Truth: On Vote Leakage in Ensemble Private Learning