arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2507.20152cs.CLcs.AI
收藏

基于对话AI的LLM用户模拟器中的目标对齐

Goal Alignment in LLM-Based User Simulators for Conversational AI

University of Illinois Urbana-Champaign(伊利诺伊大学厄巴纳-香槟分校)

Shuhaib Mehri, Xiaocheng Yang, Takyoung Kim, Gokhan Tur, Shikib Mehri, Dilek Hakkani-Tür

AI总结:

本文提出UGST框架,通过三阶段方法改进用户模拟器的目标对齐能力,并在两个基准测试中取得显著成效。

摘要:

用户模拟器对于对话AI至关重要,通过模拟交互实现可扩展的代理开发和评估。尽管当前大型语言模型(LLMs)已具备先进的用户模拟能力,但我们发现它们在多轮对话中难以一致地展示目标导向行为——这是影响其在下游应用中可靠性的关键限制。我们引入了用户目标状态跟踪(UGST),一种新的框架,用于跟踪对话中的用户目标进展。利用UGST,我们提出了一种三阶段方法,用于开发能够自主跟踪目标进展并推理以生成目标对齐响应的用户模拟器。此外,我们建立了全面的评估指标,用于衡量用户模拟器中的目标对齐程度,并证明我们的方法在两个基准(MultiWOZ 2.4和τ-Bench)上实现了显著改进。我们的贡献填补了对话AI中的关键空白,并将UGST确立为开发目标对齐用户模拟器的重要框架。

英文摘要:

User simulators are essential to conversational AI, enabling scalable agent development and evaluation through simulated interactions. While current Large Language Models (LLMs) have advanced user simulation capabilities, we reveal that they struggle to consistently demonstrate goal-oriented behavior across multi-turn conversations--a critical limitation that compromises their reliability in downstream applications. We introduce User Goal State Tracking (UGST), a novel framework that tracks user goal progression throughout conversations. Leveraging UGST, we present a three-stage methodology for developing user simulators that can autonomously track goal progression and reason to generate goal-aligned responses. Moreover, we establish comprehensive evaluation metrics for measuring goal alignment in user simulators, and demonstrate that our approach yields substantial improvements across two benchmarks (MultiWOZ 2.4 and τ-Bench). Our contributions address a critical gap in conversational AI and establish UGST as an essential framework for developing goal-aligned user simulators.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑