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

Neural metrics for machine translation evaluation, such as COMET, exhibitsignificant improvements in their correlation with human judgments, as comparedto traditional metrics based on lexical overlap, such as BLEU. Yet, neuralmetrics are, to a great extent, "black boxes" returning a single sentence-levelscore without transparency about the decision-making process. In this work, wedevelop and compare several neural explainability methods and demonstrate theireffectiveness for interpreting state-of-the-art fine-tuned neural metrics. Ourstudy reveals that these metrics leverage token-level information that can bedirectly attributed to translation errors, as assessed through comparison oftoken-level neural saliency maps with Multidimensional Quality Metrics (MQM)annotations and with synthetically-generated critical translation errors. Toease future research, we release our code at:https://github.com/Unbabel/COMET/tree/explainable-metrics.

The Inside Story: Towards Better Understanding of Machine Translation Neural Evaluation Metrics