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

Augmenting pre-trained language models with knowledge graphs (KGs) hasachieved success on various commonsense reasoning tasks. However, for a giventask instance, the KG, or certain parts of the KG, may not be useful. AlthoughKG-augmented models often use attention to focus on specific KG components, theKG is still always used, and the attention mechanism is never explicitly taughtwhich KG components should be used. Meanwhile, saliency methods can measure howmuch a KG feature (e.g., graph, node, path) influences the model to make thecorrect prediction, thus explaining which KG features are useful. This paperexplores how saliency explanations can be used to improve KG-augmented models'performance. First, we propose to create coarse (Is the KG useful?) and fine(Which nodes/paths in the KG are useful?) saliency explanations. Second, tomotivate saliency-based supervision, we analyze oracle KG-augmented modelswhich directly use saliency explanations as extra inputs for guiding theirattention. Third, we propose SalKG, a framework for KG-augmented models tolearn from coarse and/or fine saliency explanations. Given saliencyexplanations created from a task's training set, SalKG jointly trains the modelto predict the explanations, then solve the task by attending to KG featureshighlighted by the predicted explanations. On three commonsense QA benchmarks(CSQA, OBQA, CODAH) and a range of KG-augmented models, we show that SalKG canyield considerable performance gains -- up to 2.76% absolute improvement onCSQA.

SalKG: Learning From Knowledge Graph Explanations for Commonsense Reasoning