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

We present a new zero-shot dense retrieval (ZeroDR) method, COCO-DR, toimprove the generalization ability of dense retrieval by combating thedistribution shifts between source training tasks and target scenarios. Tomitigate the impact of document differences, COCO-DR continues pretraining thelanguage model on the target corpora to adapt the model to target distributionsvia COtinuous COtrastive learning. To prepare for unseen target queries,COCO-DR leverages implicit Distributionally Robust Optimization (iDRO) toreweight samples from different source query clusters for improving modelrobustness over rare queries during fine-tuning. COCO-DR achieves superioraverage performance on BEIR, the zero-shot retrieval benchmark. At BERT Basescale, COCO-DR Base outperforms other ZeroDR models with 60x larger size. AtBERT Large scale, COCO-DR Large outperforms the giant GPT-3 embedding modelwhich has 500x more parameters. Our analysis show the correlation betweenCOCO-DR's effectiveness in combating distribution shifts and improvingzero-shot accuracy. Our code and model can be found at\url{https://github.com/OpenMatch/COCO-DR}.

COCO-DR: Combating Distribution Shifts in Zero-Shot Dense Retrieval with Contrastive and Distributionally Robust Learning