SPASM：面向多轮对话生成的稳定角色驱动型智能体模拟框架

SPASM: Stable Persona-driven Agent Simulation for Multi-turn Dialogue Generation

April 10, 2026

作者: Han Luo, Guy Laban

cs.AI

摘要

大型语言模型正日益广泛应用于多轮对话场景，如课业辅导、技术支持和心理疏导等，其可靠性取决于能否在长对话跨度中保持稳定的角色设定、人物特征和对话目标。当LLM被用于生成训练与评估所需的合成对话时，这一要求尤为关键——因为LLM之间的对话会逐渐累积身份相关故障，例如人物特征漂移、角色混淆以及"回声效应"（即一方逐渐模仿对话伙伴的言行）。我们提出SPASM（基于稳定人物特征的多轮对话生成智能体模拟框架），该模块化框架以稳定性为核心，将模拟流程分解为：（i）通过模式采样、合理性验证和自然语言人物塑造实现人物创建；（ii）客户端-应答端对话生成；（iii）基于连贯性判断的终止检测。为在不调整模型权重的前提下提升长对话稳定性，我们提出自我中心语境投射技术：对话历史以视角无关的形式存储，并在生成对话前确定性地投射至每个智能体的自我中心视角。基于三种LLM骨干模型（GPT-4o-mini、DeepSeek-V3.2、Qwen-Plus）和九组客户端-应答端配对，我们构建了包含4,500组人物特征和45,000场对话的数据集（每组配对包含500组人物特征×10场对话）。消融实验表明，ECP技术显著降低了人物特征漂移，且经人工验证完全消除了回声效应；嵌入分析不仅还原了人物特征结构，更揭示了应答端主导的强交互几何模式。代码已开源：https://github.com/lhannnn/SPASM。

English

Large language models are increasingly deployed in multi-turn settings such as tutoring, support, and counseling, where reliability depends on preserving consistent roles, personas, and goals across long horizons. This requirement becomes critical when LLMs are used to generate synthetic dialogues for training and evaluation, since LLM--LLM conversations can accumulate identity-related failures such as persona drift, role confusion, and "echoing", where one agent gradually mirrors its partner. We introduce SPASM (Stable Persona-driven Agent Simulation for Multi-turn dialogue generation), a modular, stability-first framework that decomposes simulation into (i) persona creation via schema sampling, plausibility validation, and natural-language persona crafting, (ii) Client--Responder dialogue generation, and (iii) termination detection for coherent stopping. To improve long-horizon stability without changing model weights, we propose Egocentric Context Projection (ECP): dialogue history is stored in a perspective-agnostic representation and deterministically projected into each agent's egocentric view before generation. Across three LLM backbones (GPT-4o-mini, DeepSeek-V3.2, Qwen-Plus) and nine Client--Responder pairings, we construct a dataset of 4,500 personas and 45,000 conversations (500 personas X 10 conversations per pairing). Ablations show ECP substantially reduces persona drift and, under human validation, eliminates echoing; embedding analyses recover persona structure and reveal strong responder-driven interaction geometry. Our code is available at https://github.com/lhannnn/SPASM.