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

Contrastive learning has recently established itself as a powerfulself-supervised learning framework for extracting rich and versatile datarepresentations. Broadly speaking, contrastive learning relies on a dataaugmentation scheme to generate two versions of the input data and learnslow-dimensional representations by maximizing a normalized temperature-scaledcross entropy loss (NT-Xent) to identify augmented samples corresponding to thesame original entity. In this paper, we investigate the potential of deployingcontrastive learning in combination with Graph Neural Networks for embeddingnodes in a graph. Specifically, we show that the quality of the resultingembeddings and training time can be significantly improved by a simplecolumn-wise postprocessing of the embedding matrix, instead of the row-wisepostprocessing via multilayer perceptrons (MLPs) that is adopted by themajority of peer methods. This modification yields improvements in downstreamclassification tasks of up to 1.5% and even beats existing state-of-the-artapproaches on 6 out of 8 different benchmarks. We justify our choices ofpostprocessing by revisiting the "alignment vs. uniformity paradigm", and showthat column-wise post-processing improves both "alignment" and "uniformity" ofthe embeddings.

A Simplified Framework for Contrastive Learning for Node Representations