@inproceedings{wang-etal-2025-multipl,
    title = "{M}ulti{PL}-{M}o{E}: Multi-Programming-Lingual Extension of Large Language Models through Hybrid Mixture-of-Experts",
    author = "Wang, Qing  and
      Han, Xue  and
      Wang, Jiahui  and
      Xing, Lehao  and
      Hu, Qian  and
      Zhang, Lianlian  and
      Deng, Chao  and
      Feng, Junlan",
    editor = "Christodoulopoulos, Christos  and
      Chakraborty, Tanmoy  and
      Rose, Carolyn  and
      Peng, Violet",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2025",
    month = nov,
    year = "2025",
    address = "Suzhou, China",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/author-page-yu-wang-polytechnic/2025.findings-emnlp.686/",
    doi = "10.18653/v1/2025.findings-emnlp.686",
    pages = "12817--12828",
    ISBN = "979-8-89176-335-7",
    abstract = "Despite LLMs' excellent code creation capabilities, multilingual code generation remains extremely challenging. To address this, we intent to improve the multi-programming-lingual (MultiPL) performance of the base LLMs while retaining the most popular ones using restricted computational resources. We consider MultiPL to be a special case of multiple natural languages and propose a MultiPL extension of LLMs utilizing a hybrid mixture of experts (MoE), called MultiPL-MoE. Specifically, MultiPL-MoE combines two paired MoEs to optimize expert selection at both the token and segment levels. The **token-level MoE** is a standard upcycling MoE structure with a shared expert and a novel gate weight normalization approach that aids in the final fusion with the segment-level MoE. The **segment-level MoE** incorporates two innovative designs to better capture the syntactic structure and contextual patterns of programming languages: First, using a sliding window to partition the input token sequence into multiple segments; Then, adopting an expert-choice routing strategy that allows experts to select the top-k segments. The results of the experiment proved the effectiveness of MultiPL-MoE."
}