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

Decisions of complex language understanding models can be rationalized bylimiting their inputs to a relevant subsequence of the original text. Arationale should be as concise as possible without significantly degrading taskperformance, but this balance can be difficult to achieve in practice. In thispaper, we show that it is possible to better manage this trade-off byoptimizing a bound on the Information Bottleneck (IB) objective. Our fullyunsupervised approach jointly learns an explainer that predicts sparse binarymasks over sentences, and an end-task predictor that considers only theextracted rationale. Using IB, we derive a learning objective that allowsdirect control of mask sparsity levels through a tunable sparse prior.Experiments on ERASER benchmark tasks demonstrate significant gains overnorm-minimization techniques for both task performance and agreement with humanrationales. Furthermore, we find that in the semi-supervised setting, a modestamount of gold rationales (25% of training examples) closes the gap with amodel that uses the full input.

An Information Bottleneck Approach for Controlling Conciseness in Rationale Extraction