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

Has there been real progress in multi-hop question-answering? Models oftenexploit dataset artifacts to produce correct answers, without connectinginformation across multiple supporting facts. This limits our ability tomeasure true progress and defeats the purpose of building multi-hop QAdatasets. We make three contributions towards addressing this. First, weformalize such undesirable behavior as disconnected reasoning across subsets ofsupporting facts. This allows developing a model-agnostic probe for measuringhow much any model can cheat via disconnected reasoning. Second, using a notionof \emph{contrastive support sufficiency}, we introduce an automatictransformation of existing datasets that reduces the amount of disconnectedreasoning. Third, our experiments suggest that there hasn't been much progressin multi-hop QA in the reading comprehension setting. For a recent large-scalemodel (XLNet), we show that only 18 points out of its answer F1 score of 72 onHotpotQA are obtained through multifact reasoning, roughly the same as that ofa simpler RNN baseline. Our transformation substantially reduces disconnectedreasoning (19 points in answer F1). It is complementary to adversarialapproaches, yielding further reductions in conjunction.

Is Multihop QA in DiRe Condition? Measuring and Reducing Disconnected Reasoning