MUSE: Multi-Domain Chinese User Simulation via Self-Evolving Profiles and Rubric-Guided Alignment

TL;DR

MUSE is a multi-domain Chinese user simulation framework that uses self-evolving profiles and rubric-guided reinforcement learning to produce realistic, coherent, and persona-consistent responses over long interactions.

Contribution

The paper introduces MUSE, a novel multi-domain Chinese user simulator with self-evolving profiles and rubric-guided training, improving response realism and long-term consistency.

Findings

MUSE outperforms baselines in realism and coherence.

It maintains persona consistency over extended dialogues.

The framework enhances multi-turn response quality.

Abstract

User simulators are essential for the scalable training and evaluation of interactive AI systems. However, existing approaches often rely on shallow user profiling, struggle to maintain persona consistency over long interactions, and are largely limited to English or single-domain settings. We present MUSE, a multi-domain Chinese user simulation framework designed to generate human-like, controllable, and behaviorally consistent responses. First, we propose Iterative Profile Self-Evolution (IPSE), which gradually optimizes user profiles by comparing and reasoning discrepancies between simulated trajectories and real dialogue behaviors. We then apply Role-Reversal Supervised Fine-Tuning to improve local response realism and human-like expression. To enable fine-grained behavioral alignment, we further train a specialized rubric-based reward model and incorporate it into rubric-guided multi-turn reinforcement learning, which optimizes the simulator at the dialogue level and enhances long-horizon behavioral consistency. Experiments show that MUSE consistently outperforms strong baselines in both utterance-level and session-level evaluations, generating responses that are more realistic, coherent, and persona-consistent over extended interactions.