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

Contrastive learning has achieved impressive success in generation tasks tomilitate the "exposure bias" problem and discriminatively exploit the differentquality of references. Existing works mostly focus on contrastive learning onthe instance-level without discriminating the contribution of each word, whilekeywords are the gist of the text and dominant the constrained mappingrelationships. Hence, in this work, we propose a hierarchical contrastivelearning mechanism, which can unify hybrid granularities semantic meaning inthe input text. Concretely, we first propose a keyword graph via contrastivecorrelations of positive-negative pairs to iteratively polish the keywordrepresentations. Then, we construct intra-contrasts within instance-level andkeyword-level, where we assume words are sampled nodes from a sentencedistribution. Finally, to bridge the gap between independent contrast levelsand tackle the common contrast vanishing problem, we propose an inter-contrastmechanism that measures the discrepancy between contrastive keyword nodesrespectively to the instance distribution. Experiments demonstrate that ourmodel outperforms competitive baselines on paraphrasing, dialogue generation,and storytelling tasks.

Keywords and Instances: A Hierarchical Contrastive Learning Framework Unifying Hybrid Granularities for Text Generation