让聊天角色更稳定，通过边界感知学习提升一致性。
ERABAL: Enhancing Role-Playing Agents through Boundary-Aware Learning
- 设计边界感知训练框架，精准捕捉角色边界
- 用更少对话数据实现比主流方法更好的角色一致性表现
- 适合需要高角色稳定性的人机交互场景
角色扮演是人机交互领域的新应用，主要通过大语言模型（LLM）与指定角色对齐训练实现。尽管进展显著，角色扮演代理（RPLAs）在面对与角色属性相关的边界查询时，仍难以保持角色一致性。本文提出ERABAL框架，通过边界感知学习增强角色扮演能力。该框架包含生成特定角色对话的流水线及配套对齐训练方法。全面评估表明，ERABAL高效且有效：仅需比领先方法少得多的对话数据，就在WikiRoleEval、CharacterEval和MT-Bench的角色扮演子集上显著优于通用基线模型。代码与数据集将公开以支持后续研究。
原文摘要 · Abstract (English)
Role-playing is an emerging application in the field of Human-Computer Interaction (HCI), primarily implemented through the alignment training of a large language model (LLM) with assigned characters. Despite significant progress, role-playing agents (RPLAs) still struggle with maintaining role-consistency across conversations, particularly when confronted with boundary queries subtly related to character attributes. In this paper, we present ERABAL, a framework aimed at enhancing RPLAs' role-playing capabilities through boundary-aware learning. ERABAL encompasses a generation pipeline for role-specific dialogues and a concomitant methodology for alignment training. Through comprehensive evaluations, we demonstrate that ERABAL is both efficient and effective. By training with significantly fewer dialogues than those used in leading approaches, ERABAL achieves notable improvements across WikiRoleEval, CharacterEval, and the role-playing subset of MT-Bench compared to the generalist baseline models. Our code and datasets will be made publicly available to support further research.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。