aiqbench analysis
AdaMARP introduces a novel framework for LLM role-playing that significantly improves immersion by integrating dynamic scene management and multi-agent orchestration. The research demonstrates that smaller models can outperform larger commercial counterparts when equipped with structured environmental grounding and explicit action rationales, marking progress in generalizable interactive narrative generation. This work advances the state of AI agents capable of maintaining consistent character identities within evolving storylines without relying on static prompts.
AdaMARP: An Adaptive Multi-Agent Interaction Framework for General Immersive Role-Playing
arXiv:2601.11007v2 Announce Type: replace Abstract: LLM role-playing aims to portray arbitrary characters in interactive narratives, yet existing systems often suffer from limited immersion and adaptability. They typically under-model dynamic environmental information and assume largely static scenes and casts, offering insufficient support for multi-character orchestration, scene transitions, and on-the-fly character introduction. We propose an adaptive multi-agent role-playing framework, AdaMARP, featuring an immersive message format that interleaves [Thought], (Action), , and Speech, together with an explicit Scene Manager that governs role-playing through discrete actions (init_scene, pick_speaker, switch_scene, add_role, end) accompanied by rationales. To train these capabilities, we construct AdaRPSet for the Actor Model and AdaSMSet for supervising orchestration decisions, and introduce AdaptiveBench for trajectory-level evaluation. Experiments across multiple backbones and model scales demonstrate consistent improvements: AdaRPSet enhances character consistency, environment grounding, and narrative coherence, with an 8B actor outperforming several commercial LLMs, while AdaSMSet enables smoother scene transitions and more natural role introductions, surpassing Claude Sonnet 4.5 using only a 14B LLM.
Abstract:LLM role-playing aims to portray arbitrary characters in interactive narratives, yet existing systems often suffer from limited immersion and adaptability. They typically under-model dynamic environmental information and assume largely static scenes and casts, offering insufficient support for multi-character orchestration, scene transitions, and on-the-fly character introduction. We propose an adaptive multi-agent role-playing framework, AdaMARP, featuring an immersive message format that interleaves [Thought], (Action), <Environment>, and Speech, together with an explicit Scene Manager that governs role-playing through discrete actions (init_scene, pick_speaker, switch_scene, add_role, end) accompanied by rationales. To train these capabilities, we construct AdaRPSet for the Actor Model and AdaSMSet for supervising orchestration decisions, and introduce AdaptiveBench for trajectory-level evaluation. Experiments across multiple backbones and model scales demonstrate consistent improvements: AdaRPSet enhances character consistency, environment grounding, and narrative coherence, with an 8B actor outperforming several commercial LLMs, while AdaSMSet enables smoother scene transitions and more natural role introductions, surpassing Claude Sonnet 4.5 using only a 14B LLM.
Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
arXiv:2601.11007 [cs.AI]
(or arXiv:2601.11007v2 [cs.AI] for this version)
https://doi.org/10.48550/arXiv.2601.11007 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 251 words
Related coverage
The Machines that Make the Machines
NVIDIA · Oct 7, 2026
Amazon Research Awards recipients announced
Amazon Science · May 27, 2026
Meta-Multi-Agent Reinforcement Learning for Fast Adaptation of Interactive Policies with Applications to Autonomous Driving
arXiv · Oct 2, 2026
LLM-Based Synthetic Ground Truth Generation for Audio-Based Emotion Classification via In-Context Learning
arXiv · Jun 16, 2026