@inproceedings{liu-etal-2026-papermentor,
    title = "{P}aper{M}entor: A Human-Centered Multi-Agent Writing Tutor for {AI} Research Papers in Overleaf",
    author = {Liu, Jiarui  and
      Zhang, Terry Jingchen  and
      Faulkner, Ryan  and
      Huang, Xuanqiang Angelo  and
      Zouhar, Vil{\'e}m  and
      Glandorf, Dominik  and
      Dahlgren, Isabel  and
      Dagli, Rishit  and
      Chen, Yuen  and
      Leeb, Felix  and
      Truong, Van Q.  and
      Pandey, Punya Syon  and
      Bicker, Yves  and
      Majumder, Suvajit  and
      Jiang, Wenyuan  and
      Qiu, Zeju  and
      Pal Chowdhury, Sankalan  and
      Sachan, Mrinmaya  and
      Sch{\"o}lkopf, Bernhard  and
      Diab, Mona T.  and
      Jin, Zhijing},
    editor = "Durrett, Greg  and
      Jian, Ping",
    booktitle = "Proceedings of the 64th Annual Meeting of the {A}ssociation for {C}omputational {L}inguistics (Volume 3: System Demonstrations)",
    month = jul,
    year = "2026",
    address = "San Diego, California, United States",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-acl/2026.acl-demo.39/",
    pages = "397--407",
    ISBN = "979-8-89176-392-0",
    abstract = "Expert writing feedback from experienced researchers is critical for early-career scholars to improve their manuscripts, yet high-quality feedback often remains scarce because reviewing research papers is labor-intensive. Emerging AI-powered writing assistants largely focus on grammar fixes or simulating peer review with final scores, yet they fall short of providing concrete, actionable suggestions that help students improve their papers during drafting. We present PaperMentor, a human-centered writing assistant system that delivers actionable suggestions as Overleaf-native inline comments while leaving the actual writing entirely to human authors. PaperMentor integrates an expert skill library carefully curated from established researchers' writing advice with 12 specialized agents covering different aspects of paper writing, such as formatting compliance, phrasing accuracy, and terminology consistency. In a user study (n=14), 90.6{\%} of the generated comments were rated actionable and 67.5{\%} were rated valid, significantly outperforming a GPT-5.2 baseline without the skill library. We release PaperMentor as open source for public use."
}