@inproceedings{liu-etal-2026-dialogue,
    title = "Dialogue is Better Than Monologue: Instructing Meidcal {LLM}s via Strategic Conversations",
    author = "Liu, Zijie  and
      Zhao, Xinyu  and
      Peng, Jie  and
      Duan, Jinhao  and
      Zhu, Zhuangdi  and
      Chen, Qingyu  and
      Xu, Kaidi  and
      Hu, Xia  and
      Chen, Tianlong",
    editor = "Demberg, Vera  and
      Inui, Kentaro  and
      Marquez, Llu{\'i}s",
    booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {EACL} 2026",
    month = mar,
    year = "2026",
    address = "Rabat, Morocco",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/more-markup/2026.findings-eacl.149/",
    doi = "10.18653/v1/2026.findings-eacl.149",
    pages = "2858--2872",
    ISBN = "979-8-89176-386-9",
    abstract = "In real clinical practice, clinicians must sift through noisy and often conflicting information, progressively gathering and sequencing evidence before reaching conclusions. However, existing tuning methods for medical AI models are typically \textbf{monologue-based} {---} that is, models are fine-tuned on static question answering (QA) tasks or medical articles, which fail to reflect the interactive and iterative nature of clinical reasoning. To bridge this gap, we introduce \textbf{MuddyMaze}, a benchmark designed to expose the limitations of current \textbf{monologue-based} tuning, and construct a large dialogue dataset of 22.2k doctor{--}patient interactions that capture stepwise diagnostic reasoning validated by medical experts. Building on those, we propose \textbf{dialogue-tuning}, a new fine-tuning paradigm that captures the internal reasoning dynamics unfolding across interactions.To assess the effectiveness of our approach, we evaluated \textit{dialogue-tuned} models on \textbf{MuddyMaze}, where they outperform \textit{monologue-tuned} baselines (e.g., MedQA) by +16.1{\%} in one-round and +4.1{\%} in multi-round evidence ranking, while maintaining or even improving accuracy on standard medical QA benchmarks (e.g., PubMedQA). These results indicate that \textbf{dialogue-tuning} not only enhances reasoning robustness and evidence integration but also preserves the factual precision of traditional QA performance."
}