发表机构
University of Göttingen（哥廷根大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本研究针对LLM代理在在线会议中无法识别发言时机的问题，提出CAPA架构，经实验可大幅降低沉默率、提升可信恢复率且保持低幻觉率，证明会议状态是缩小识别差距的关键。
AI 中文摘要
在在线会议代理场景中，大语言模型（LLM）智能体无法识别发言时机，由于缺乏跟踪立场、覆盖范围和发言权的结构化方式，它们会错过应做出贡献的时刻。仅使用提示词的代理在AMI语料库中，对缺席参与者的51.4%的发言机会保持沉默。我们提出CAPA（Collaborative Agent Predictive Architecture，协作智能体预测架构），一种用于在线会议代理的架构：感知器（Perceiver）从每个观察到的对话轮次更新会议状态；预测器（Predictor）预测对话的后续走向；控制器（Controller）决定是否发言以及提出哪个主张；生成器（Generator）以参与者的风格表述选定的贡献；两名评判员（judges）根据下一个观察到的轮次对预测和行动进行评分；重新校准器（Recalibrator）根据这些评判更新会议状态，以供未来决策。为评估在线代理，我们引入了基于对话片段（episode）的协议，该协议围绕参与者的实际思想单元对代理是否贡献、何时贡献以及贡献内容进行评分，该协议的受模式约束的LLM评判员与人工标注的Cohen's kappa值为0.71。在137场AMI会议上，CAPA将沉默率从51.4%降至2.5%，使可信恢复率翻倍（从26.1升至52.2），同时将幻觉率保持在0.6%。失败模式从遗漏转向选择，每个剩余的接近失误都可归因于架构的特定模块。机制消融实验表明，会议状态是缩小识别差距的关键因素，仅靠原始上下文缩放无法实现这一点。
英文摘要
In online meeting delegation, LLM agents fail to recognize when to speak. With no structured way to track stances, coverage, and floor, they miss the moments where they should contribute. Prompt-only delegates stay silent on 51.4% of the absent participant's talking opportunities on the AMI corpus. We present CAPA (Collaborative Agent Predictive Architecture), an architecture for online meeting delegation. A Perceiver updates the meeting state from each observed turn. A Predictor forecasts how the conversation will continue. A Controller decides whether to speak and which proposition to surface. A Generator phrases the chosen contribution in the participant's style. Two judges score the forecast and the action against the next observed turn. A Recalibrator updates the meeting state from those verdicts for future decisions. To evaluate online delegation, we introduce an episode-level protocol that scores whether, when, and what a delegate contributes around the participant's actual idea units. The protocol's schema-constrained LLM judges align with human annotations at Cohen's kappa = 0.71. On 137 AMI meetings, CAPA reduces the silence rate from 51.4% to 2.5%, doubles credited recovery (26.1 --> 52.2), and keeps hallucination at 0.6%. The failure mode shifts from omission to selection, with each residual near-miss attributable to a specific module of the architecture. Mechanism ablations identify the meeting state as the lever that closes the recognition gap, where raw-context scaling alone does not.
CommentsAccepted at EMNLP 2026 Main
