AdaMARP: An Adaptive Multi-Agent Interaction Framework for General Immersive Role-Playing

TL;DR

AdaMARP introduces an adaptive multi-agent framework for immersive role-playing with improved character consistency, scene management, and narrative coherence, leveraging new training sets and evaluation benchmarks.

Contribution

The paper presents AdaMARP, a novel framework with an immersive message format and explicit scene management, enhancing multi-character orchestration and adaptability in role-playing systems.

Findings

AdaRPSet improves character consistency and narrative coherence.

AdaSMSet enables smoother scene transitions and natural role introductions.

An 8B model outperforms several commercial LLMs in role-playing tasks.

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.