Saved in:
Bibliographic Details
Main Authors: Yu, Seunguk, Kim, Kyeonghyun, Yun, Jungmin, Kim, Youngbin
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2507.03378
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908434145738752
author Yu, Seunguk
Kim, Kyeonghyun
Yun, Jungmin
Kim, Youngbin
author_facet Yu, Seunguk
Kim, Kyeonghyun
Yun, Jungmin
Kim, Youngbin
contents Although LLMs have made significant progress in various languages, there are still concerns about their effectiveness with low-resource agglutinative languages compared to languages such as English. In this study, we focused on Korean, a language known for its complex sentence endings, and evaluated LLMs on this challenging aspect. We introduce the Korean Sentence Endings (KoSEnd) dataset, which includes 3,000 sentences, each annotated for the naturalness of 15 sentence ending forms. These were collected from diverse sources to cover a range of contexts. We evaluated 11 LLMs to assess their understanding of Korean sentence endings, analyzing them based on parameter count and prediction consistency. Notably, we found that informing models about the possibility of missing sentence endings improved performance, highlighting the impact of explicitly considering certain linguistic features.
format Preprint
id arxiv_https___arxiv_org_abs_2507_03378
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Making Sense of Korean Sentences: A Comprehensive Evaluation of LLMs through KoSEnd Dataset
Yu, Seunguk
Kim, Kyeonghyun
Yun, Jungmin
Kim, Youngbin
Computation and Language
Although LLMs have made significant progress in various languages, there are still concerns about their effectiveness with low-resource agglutinative languages compared to languages such as English. In this study, we focused on Korean, a language known for its complex sentence endings, and evaluated LLMs on this challenging aspect. We introduce the Korean Sentence Endings (KoSEnd) dataset, which includes 3,000 sentences, each annotated for the naturalness of 15 sentence ending forms. These were collected from diverse sources to cover a range of contexts. We evaluated 11 LLMs to assess their understanding of Korean sentence endings, analyzing them based on parameter count and prediction consistency. Notably, we found that informing models about the possibility of missing sentence endings improved performance, highlighting the impact of explicitly considering certain linguistic features.
title Making Sense of Korean Sentences: A Comprehensive Evaluation of LLMs through KoSEnd Dataset
topic Computation and Language
url https://arxiv.org/abs/2507.03378