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

Deep neural networks are vulnerable to adversarial attacks, where a smallperturbation to an input alters the model prediction. In many cases, maliciousinputs intentionally crafted for one model can fool another model. In thispaper, we present the first study to systematically investigate thetransferability of adversarial examples for text classification models andexplore how various factors, including network architecture, tokenizationscheme, word embedding, and model capacity, affect the transferability ofadversarial examples. Based on these studies, we propose a genetic algorithm tofind an ensemble of models that can be used to induce adversarial examples tofool almost all existing models. Such adversarial examples reflect the defectsof the learning process and the data bias in the training set. Finally, wederive word replacement rules that can be used for model diagnostics from theseadversarial examples.

On the Transferability of Adversarial Attacksagainst Neural Text Classifier