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

Current Natural Language Inference (NLI) models achieve impressive results,sometimes outperforming humans when evaluating on in-distribution test sets.However, as these models are known to learn from annotation artefacts anddataset biases, it is unclear to what extent the models are learning the taskof NLI instead of learning from shallow heuristics in their training data. Weaddress this issue by introducing a logical reasoning framework for NLI,creating highly transparent model decisions that are based on logical rules.Unlike prior work, we show that improved interpretability can be achievedwithout decreasing the predictive accuracy. We almost fully retain performanceon SNLI, while also identifying the exact hypothesis spans that are responsiblefor each model prediction. Using the e-SNLI human explanations, we verify thatour model makes sensible decisions at a span level, despite not using any spanlabels during training. We can further improve model performance and span-leveldecisions by using the e-SNLI explanations during training. Finally, our modelis more robust in a reduced data setting. When training with only 1,000examples, out-of-distribution performance improves on the MNLI matched andmismatched validation sets by 13% and 16% relative to the baseline. Trainingwith fewer observations yields further improvements, both in-distribution andout-of-distribution.

Logical Reasoning with Span-Level Predictions for Interpretable and Robust NLI Models