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

Generative question answering (QA) models generate answers to questionseither solely based on the parameters of the model (the closed-book setting) oradditionally retrieving relevant evidence (the open-book setting). GenerativeQA models can answer some relatively complex questions, but the mechanismthrough which they do so is still poorly understood. We perform several studiesaimed at better understanding the multi-hop reasoning capabilities ofgenerative QA models. First, we decompose multi-hop questions into multiplecorresponding single-hop questions, and find marked inconsistency in QA models'answers on these pairs of ostensibly identical question chains. Second, we findthat models lack zero-shot multi-hop reasoning ability: when trained only onsingle-hop questions, models generalize poorly to multi-hop questions. Finally,we demonstrate that it is possible to improve models' zero-shot multi-hopreasoning capacity through two methods that approximate real multi-hop naturallanguage (NL) questions by training on either concatenation of single-hopquestions or logical forms (SPARQL). In sum, these results demonstrate thatmulti-hop reasoning does not emerge naturally in generative QA models, but canbe encouraged by advances in training or modeling techniques.

Understanding and Improving Zero-shot Multi-hop Reasoning in Generative Question Answering