GPT-4-turbo dataset.

<div><p>This study examines the multilingual capabilities of GPT, focusing on its handling of syntactic ambiguity across English, Korean, and Japanese. We investigate whether GPT can capture language-specific attachment preferences or if it relies primarily on English-centric training pa...

Full description

Saved in:

Bibliographic Details
Main Author:	Myung Hye Yoo (21672909) (author)
Other Authors:	Joungmin Kim (21672912) (author), Sanghoun Song (14666774) (author)
Published:	2025
Subjects:	Evolutionary Biology Sociology Cancer Science Policy Environmental Sciences not elsewhere classified Biological Sciences not elsewhere classified specific attachment preferences overall results suggest lacks clear preferences bias toward high like language processing centric training patterns investigate whether gpt gpt &# 8217 mini &# 8212 centric generalization capture language xlink "> testing ground relies primarily preferred interpretations performance aligns nuanced engagement multilingual proficiency multilingual capabilities model size likely reflecting findings underscore findings reveal although differences
Tags:	Add Tag No Tags, Be the first to tag this record!

Be the first to leave a comment!

GPT-4-turbo dataset.

Similar Items