Small Language Models Can Learn Linguistic Representations from Character-Level Inputs
Small language models trained on character-level inputs can capture linguistic structures at various levels, including syntax, lexicon, and phonetics, performing comparably to or even outperforming larger subword-based models.
Comprehensive Exploration of Korean Verb Lexicon: A User-Friendly Interface and Python Library for Accessing Subcategorization Frames
This paper introduces a user-friendly web interface and a Python library to facilitate easy access and manipulation of the extensive linguistic information in the Sejong dictionary, with a focus on Korean verb subcategorization frames.
Exploring Coping Strategies in Emotional Responses: A Role-Playing Corpus for Computational Analysis
Emotions are fundamentally linked to coping strategies that people use to deal with salient situations. This study introduces a novel corpus, COPING, to investigate how these coping strategies (attack, contact, distance, reject) are expressed in language and can be computationally identified.
Mitigating Bias in Large Language Model Evaluation: A Systematic Approach
Existing LLM evaluators suffer from bias towards superficial quality, overlooking instruction following ability. This work proposes systematic methods to mitigate the bias, including online calibration and offline contrastive training, effectively improving the fairness of LLM evaluation.
Benchmarking Large Language Models' Alignment with Human Language Use Patterns
Large language models (LLMs) can generate human-like text, but the extent to which they truly replicate human language use patterns remains unclear. This study introduces a comprehensive psycholinguistic benchmark to systematically evaluate the humanlikeness of 20 prominent LLMs across various linguistic levels.
언어 현장 조사를 안내하는 신경망 모델: 형태론적 굴절에 대한 사례 연구
언어 현장 조사 과정에서 신경망 모델을 활용하여 데이터 수집의 효율성을 높일 수 있다.
Leveraging Neural Models to Enhance Linguistic Fieldwork Efficiency: A Case Study on Morphological Inflection
Neural models can guide linguists during fieldwork by optimizing the data collection process and accounting for the dynamics of linguist-speaker interactions.
Structural Priming in Language Models: Insights from Lexical and Frequency-Based Factors
Language models exhibit structural priming effects that can be explained by inverse frequency effects, such as prime surprisal and verb preference, as well as lexical dependence between prime and target.
Computational Analysis of Character Representation Shifts Across Online Fan Communities: A Case Study of Harry Potter
Online fan communities, such as fanfiction and forums, collaboratively reconstruct and renegotiate narrative elements like characters, leading to divergent representations from the original source material.
Sense-Specific Pitch Contours of Monosyllabic Words in Conversational Taiwan Mandarin
The realization of pitch contours of monosyllabic words in spontaneous Taiwan Mandarin speech is co-determined by tonal context, word identity, and word sense.
© 2024 by Linnk AI