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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

We investigate MT evaluation metric performance on adversarially-synthesizedtexts, to shed light on metric robustness. We experiment with word- andcharacter-level attacks on three popular machine translation metrics:BERTScore, BLEURT, and COMET. Our human experiments validate that automaticmetrics tend to overpenalize adversarially-degraded translations. We alsoidentify inconsistencies in BERTScore ratings, where it judges the originalsentence and the adversarially-degraded one as similar, while judging thedegraded translation as notably worse than the original with respect to thereference. We identify patterns of brittleness that motivate more robust metricdevelopment.

Robustness Tests for Automatic Machine Translation Metrics with Adversarial Attacks