arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2606.03032cs.CL
收藏

深思幻觉:诊断多智能体LLM讨论中的事实损耗与立场同质化

The Deliberative Illusion: Diagnosing Factual Attrition and Stance Homogenization in Multi-Agent LLM Deliberation

Xi’an Jiaotong University(西安交通大学);National University of Singapore(新加坡国立大学);Yunnan University(云南大学);Agency for Science, Technology and Research (A*STAR), Singapore(新加坡科技研究局(A*STAR))

Herun Wan, Jiaying Wu, Minnan Luo, Fanxiao Li, Ningnan Wang, Nancy F. Chen, Min-Yen Kan

AI总结:

本文提出DelibTrace框架,通过分解原子事实并追踪其存留,发现多智能体LLM讨论中高达72%的关键事实丢失,导致立场同质化,揭示了“同意越多、知道越少”的风险。

摘要:

多智能体LLM系统常将共识视为成功交互的证据。然而,对于深思性问题,可靠性取决于智能体是否保留了理解问题所需的事实和观点。我们识别出深思幻觉:讨论导致(1)事实损耗,即问题关键事实的逐渐丢失,以及(2)立场同质化,即不同立场向共识的崩溃。为了衡量这一过程,我们引入了DelibTrace框架,该框架将每个问题分解为原子事实,标记关键事实,将其分布在智能体之间,并追踪它们在讨论轮次中的存留。在涉及三个代表性LLM家族的伦理和新闻类深思中,多智能体讨论抹去了高达72%的关键事实。这种损失是严重的:保留的证据可能误导性地重构问题,最终立场仍锚定于基础模型先验,且单个恶意智能体可将错误信息注入不断缩小的共享上下文。这些结果揭示了一个更尖锐的风险:智能体可以在知道更少的情况下达成更多一致。我们呼吁进行评估,衡量哪些事实、不确定性和合理分歧在交互中得以存留。

英文摘要:

Multi-agent LLM systems often treat consensus as evidence of successful interaction. For deliberative problems, however, reliability depends on whether agents preserve the facts and viewpoints needed to interpret an issue. We identify the deliberative illusion: discussion produces (1) factual attrition, the progressive loss of issue-critical facts, alongside (2) stance homogenization, the collapse of diverse positions toward consensus. To measure this process, we introduce DelibTrace, a framework that decomposes each issue into atomic facts, labels issue-critical ones, distributes them across agents, and tracks their survival across discussion rounds. Across ethical and news-based deliberation with three representative LLM families, multi-agent discussion erases up to 72% of issue-critical facts. This loss is consequential: retained evidence can reconstruct the issue misleadingly, final stances remain anchored in base-model priors, and a single malicious agent can inject misinformation into the shrinking shared context. These results reveal a sharper risk: agents can agree more while knowing less. We call for evaluations that measure which facts, uncertainties, and legitimate disagreements survive interaction.

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。

↑