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/

Zendy Plus

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

Zendy Tools

Zendy Open

The state of the art on many NLP tasks is currently achieved by largepre-trained language models, which require a considerable amount ofcomputation. We explore a setting where many different predictions are made ona single piece of text. In that case, some of the computational cost duringinference can be amortized over the different tasks using a shared textencoder. We compare approaches for training such an encoder and show thatencoders pre-trained over multiple tasks generalize well to unseen tasks. Wealso compare ways of extracting fixed- and limited-size representations fromthis encoder, including different ways of pooling features extracted frommultiple layers or positions. Our best approach compares favorably to knowledgedistillation, achieving higher accuracy and lower computational cost once thesystem is handling around 7 tasks. Further, we show that through binaryquantization, we can reduce the size of the extracted representations by afactor of 16 making it feasible to store them for later use. The resultingmethod offers a compelling solution for using large-scale pre-trained models ata fraction of the computational cost when multiple tasks are performed on thesame text.

General Purpose Text Embeddings from Pre-trained Language Models for Scalable Inference