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

Despite their promising performance across various natural languageprocessing (NLP) tasks, current NLP systems are vulnerable to textualadversarial attacks. To defend against these attacks, most existing methodsapply adversarial training by incorporating adversarial examples. However,these methods have to rely on ground-truth labels to generate adversarialexamples, rendering it impractical for large-scale model pre-training which iscommonly used nowadays for NLP and many other tasks. In this paper, we proposea novel learning framework called SCAT (Self-supervised Contrastive Learningvia Adversarial Training), which can learn robust representations withoutrequiring labeled data. Specifically, SCAT modifies random augmentations of thedata in a fully labelfree manner to generate adversarial examples. Adversarialtraining is achieved by minimizing the contrastive loss between theaugmentations and their adversarial counterparts. We evaluate SCAT on two textclassification datasets using two state-of-the-art attack schemes proposedrecently. Our results show that SCAT can not only train robust language modelsfrom scratch, but it can also significantly improve the robustness of existingpre-trained language models. Moreover, to demonstrate its flexibility, we showthat SCAT can also be combined with supervised adversarial training to furtherenhance model robustness.

SCAT: Robust Self-supervised Contrastive Learning via Adversarial Training for Text Classification