对齐的错觉：检测协同对话中的隐藏分歧

中文摘要 AI 辅助

协同对话可能以表面一致告终，但参与者在目标、假设或执行计划上仍存在差异，从而形成“对齐的错觉（IoA）”。一项涵盖18场会议的真人研究证实，IoA在人类协作中频繁出现。然而IoA构成了一个悖论：若参与者意识到此类分歧，它们本就已明确表达；若未意识到，被询问时也无法阐明，致使IoA对参与者和观察者均不可见。本研究通过生成诊断性多项选择题，使IoA可被检测，参与者对这些问题的不同回答为隐藏分歧提供了直接行为证据。我们构建了IoA-Suite——一个用于检测隐藏分歧的数据集与评估协议，涵盖5种任务类型和6个领域。研究发现，即便最优模型也仅达到49.5%的F1值，瓶颈源于对话未呈现的私人上下文。随后我们基于IoA-Suite训练了IoA-Prober-8B，在IoA-Suite上达到51.8%的F1值。在上述18场真人会议中，它每场会议可揭示2.89个参与者确认未表达的隐藏分歧，且可迁移至实时人类对话。此外，在多智能体协作中，将IoA-Prober-8B与大语言模型智能体配对，可提升BigCodeBench-Hard和HiddenBench上的下游任务性能。

英文摘要

Collaborative dialogue can end with apparent agreement while participants still differ on goals, assumptions, or execution plans, creating an \textbf{illusion of alignment (IoA)}. A real-user study across 18 meetings confirms that IoA arises routinely in human collaboration. Yet IoA poses a paradox: if participants were aware of such disagreements, they would already be explicit; if not, they cannot articulate them when asked, leaving IoA invisible to both participants and observers. In this work, we make IoA detectable by generating diagnostic multiple-choice questions whose divergent answers across participants provide direct behavioral evidence of hidden disagreement. We construct \textbf{IoA-Suite}, a dataset and evaluation protocol for detecting hidden disagreement, spanning five task types and six domains. We find that even the best model attains only 49.5\% F1, with the bottleneck traced to private context that the dialogue does not surface. We then train \textbf{IoA-Prober-8B} based on IoA-Suite, reaching 51.8\% F1 on IoA-Suite. Across the aforementioned 18 real meetings, it surfaces 2.89 hidden disagreements per meeting that participants confirm they had not voiced, transferring to live human dialogue. Further, in multi-agent collaboration, pairing IoA-Prober-8B with LLM agents improves downstream task performance on BigCodeBench-Hard and HiddenBench.

发表机构

• College of AI, Tsinghua University（清华大学人工智能学院）
• Institute for AI Industry Research (AIR), Tsinghua University（清华大学人工智能产业研究院）
• Institute for AI, Tsinghua University（清华大学人工智能研究院）

机构由 AI 辅助整理，请以论文原文为准。