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

We investigate conditions under which test statistics exist that can reliablydetect examples, which have been adversarially manipulated in a white-boxattack. These statistics can be easily computed and calibrated by randomlycorrupting inputs. They exploit certain anomalies that adversarial attacksintroduce, in particular if they follow the paradigm of choosing perturbationsoptimally under p-norm constraints. Access to the log-odds is the onlyrequirement to defend models. We justify our approach empirically, but alsoprovide conditions under which detectability via the suggested test statisticsis guaranteed to be effective. In our experiments, we show that it is evenpossible to correct test time predictions for adversarial attacks with highaccuracy.

The Odds are Odd: A Statistical Test for Detecting Adversarial Examples