@inproceedings{gao-etal-2025-llms,
    title = "Can {LLM}s Simulate {L}2-{E}nglish Dialogue? An Information-Theoretic Analysis of {L}1-Dependent Biases",
    author = "Gao, Rena Wei  and
      Wu, Xuetong  and
      Kuribayashi, Tatsuki  and
      Ye, Mingrui  and
      Qi, Siya  and
      Roever, Carsten  and
      Liu, Yuanxing  and
      Yuan, Zheng  and
      Lau, Jey Han",
    editor = "Che, Wanxiang  and
      Nabende, Joyce  and
      Shutova, Ekaterina  and
      Pilehvar, Mohammad Taher",
    booktitle = "Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)",
    month = jul,
    year = "2025",
    address = "Vienna, Austria",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingestion-acl-25/2025.acl-long.219/",
    pages = "4355--4379",
    ISBN = "979-8-89176-251-0",
    abstract = "This study evaluates Large Language Models' (LLMs) ability to simulate non-native-like English use observed in human second language (L2) learners interfered with by their native first language (L1). In dialogue-based interviews, we prompt LLMs to mimic L2 English learners with specific L1s (e.g., Japanese, Thai, Urdu) across seven languages, comparing their outputs to real L2 learner data. Our analysis examines L1-driven linguistic biases, such as reference word usage and avoidance behaviors, using information-theoretic and distributional density measures. Results show that modern LLMs (e.g., Qwen2.5, LLAMA3, DeepseekV3, GPT 4o) replicate L1-dependent patterns observed in human L2 data, with distinct influences from various languages (e.g., Japanese, Korean, and Mandarin significantly affect tense agreement, and Urdu influences noun-verb collocations). Our results reveal LLMs' potential for L2 dialogue generation and evaluation for future educational applications."
}