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

Exemplar based accounts are often considered to be in direct opposition topure linguistic abstraction in explaining language learners' ability togeneralize to novel expressions. However, the recent success of neural networklanguage models on linguistically sensitive tasks suggests that perhapsabstractions can arise via the encoding of exemplars. We provide empiricalevidence for this claim by adapting an existing experiment that studies how anLM (BERT) generalizes the usage of novel tokens that belong to lexicalcategories such as Noun/Verb/Adjective/Adverb from exposure to only a singleinstance of their usage. We analyze the representational behavior of the noveltokens in these experiments, and find that BERT's capacity to generalize tounseen expressions involving the use of these novel tokens constitutes themovement of novel token representations towards regions of known categoryexemplars in two-dimensional space. Our results suggest that learners' encodingof exemplars can indeed give rise to abstraction like behavior.

Abstraction via exemplars? A representational case study on lexical category inference in BERT