@inproceedings{xu-etal-2026-adamarp,
    title = "{A}da{MARP}: An Adaptive Multi-Agent Interaction Framework for General Immersive Role-Playing",
    author = "Xu, Zhenhua  and
      Chen, Dongsheng  and
      Wang, Shuo  and
      Li, Jian  and
      Wang, Chengjie  and
      Han, Meng  and
      Wang, Yabiao",
    editor = "Liakata, Maria  and
      Moreira, Viviane P.  and
      Zhang, Jiajun  and
      Jurgens, David",
    booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {ACL} 2026",
    month = jul,
    year = "2026",
    address = "San Diego, California, United States",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-acl/2026.findings-acl.1563/",
    pages = "31230--31283",
    ISBN = "979-8-89176-395-1",
    abstract = "LLM role-playing seeks to portray arbitrary characters in interactive narratives, yet existing systems often lack immersion and adapt ability. They typically under-model dynamic environment information and assume a largely static scene/cast, offering limited support for multi-character orchestration, scene transitions, and on-the-fly character introduction. We propose an adaptive multi-agent interaction framework dubbed AdaMARP, which featuring an immersive message format that interleaves [Thought], (Action), Environment, and Speech, and an explicit Scene Manager that controls role-playing via discrete actions (init{\_}scene, pick{\_}speaker, switch{\_}scene, add{\_}role, end) with rationales. To train these abilities, we construct AdaRPSet for the Actor Model and AdaSMSet for supervising or chestration decisions, and introduce AdaptiveBench for trajectory-level evaluation. Experiments across multiple backbones and scales show consistent gains: AdaRPSet improves character consistency, environment grounding, and narrative coherence{---}an 8B actor outperforming several commercial LLMs, while AdaSMSet enables smoother scene transitions and more natural role introductions, surpassing Claude Sonnet 4.5 with only 14B LLMs."
}