智能体
arxiv.org ·
延迟验证会破坏多智能体 LLM 信念：失稳阈值与最优校正器放置
该 arXiv 论文（v2）研究多智能体 LLM 系统中验证延迟导致虚假声明在网络中传播的问题，将其建模为带 grounded corrector 节点的图上延迟共识。通过 grounded Laplacian 谱分解，作者给出验证剂量的闭式稳定性阈值：校正过强或过延迟会使共识转为振荡，通信与验证延迟重合时最不稳定，延迟为 2 时阈值为黄金分割倒数。框架还给出超模放置目标与贪心 (1-1/e) 近似，用于在有限校正器预算下选择有影响力的节点。在五个开源模型上的实验证实了预测的剂量-延迟振荡；而 grounded factual answering 使真值成为吸收边界并消除该效应，表明失稳特定于 signed-belief 任务，grounded 验证仍具稳定作用。
AI推荐理由 arXiv 预印本，作者 Igor Itkin，v2 修订版。核心新信息：将多智能体 LLM 中延迟验证导致的错误传播建模为带 grounded corrector 节点的图上延迟共识，用 grounded Laplacian 谱分解给出验证剂量的闭式稳定性阈值，指出校正过强或过延迟会把共识变为振荡，通信与验证延迟重合时最不稳定（延迟为 2 时阈值为黄金分割倒数）；并给出超模放置目标与贪心 (1-1/e) 近似来分配有限校正器预算。实验在五个开源模型上验证了剂量-延迟振荡，且 grounded factual answering 使真值成为吸收边界从而消除该效应，说明失稳特定于 signed-belief 任务。局限：仅摘要级证据，未提供实验规模、模型清单与复现细节，结论需进一步验证。
78
正文
Computer Science > Multiagent Systems
arXiv:2606.27409 (cs)
[Submitted on 25 Jun 2026 (v1), last revised 16 Sep 2026 (this version, v2)]
Title:Delayed Verification Destabilizes Multi-Agent LLM Belief: Instability Thresholds and Optimal Corrector Placement
Authors:Igor Itkin
View a PDF of the paper titled Delayed Verification Destabilizes Multi-Agent LLM Belief: Instability Thresholds and Optimal Corrector Placement, by Igor Itkin
View PDF
HTML (experimental)
Abstract:Multi-agent large language model (LLM) systems often rely on verifier and critic agents to suppress hallucinations, but verification is delayed. During this delay, false claims can propagate through the agent network. We model this process as delayed consensus on a graph with grounded corrector nodes. Spectral decomposition by the grounded Laplacian yields a closed-form stability threshold for the verification dose: correction that is too strong or too delayed can turn consensus into oscillation. The most unstable regime occurs when the communication and verification delays coincide; for delay two, the threshold is the inverse golden ratio. The same framework gives a supermodular placement objective and a greedy (1-1/e)-approximation rule for assigning a limited corrector budget to influential nodes. Experiments across five open models confirm the predicted dose-delay oscillations. By contrast, grounded factual answering makes truth an absorbing boundary and eliminates the effect, suggesting that the instability is specific to signed-belief tasks while grounded verification remains stabilizing
Comments:
Subjects:
Multiagent Systems (cs.MA); Computation and Language (cs.CL); Machine Learning (cs.LG); Systems and Control (eess.SY)
MSC classes:
93A16, 93D05, 93C55, 90C27
ACM classes:
I.2.11; I.2.7
Cite as:
arXiv:2606.27409 [cs.MA]
(or
arXiv:2606.27409v2 [cs.MA] for this version)
https://doi.org/10.48550/arXiv.2606.27409
Focus to learn more
arXiv-issued DOI via DataCite
Submission history
From: Igor Itkin [view email]
[v1]
Thu, 25 Jun 2026 10:52:54 UTC (564 KB)
[v2]
Wed, 16 Sep 2026 14:51:47 UTC (574 KB)
Full-text links:
Access Paper:
View a PDF of the paper titled Delayed Verification Destabilizes Multi-Agent LLM Belief: Instability Thresholds and Optimal Corrector Placement, by Igor Itkin
View PDF
HTML (experimental)
TeX Source
view license
Current browse context:
cs.MA
new
|
recent
| 2026-06
Change to browse by:
cs
cs.CL
cs.LG
cs.SY
eess
eess.SY
References & Citations
NASA ADS
Google Scholar
Semantic Scholar
export BibTeX citation
Loading...
BibTeX formatted citation
×
loading...
Data provided by:
Bookmark
Bibliographic Tools
Bibliographic and Citation Tools
Bibliographic Explorer Toggle
Bibliographic Explorer (What is the Explorer?)
Connected Papers Toggle
Connected Papers (What is Connected Papers?)
Litmaps Toggle
Litmaps (What is Litmaps?)
scite.ai Toggle
scite Smart Citations (What are Smart Citations?)
Code, Data, Media
Code, Data and Media Associated with this Article
alphaXiv Toggle
alphaXiv (What is alphaXiv?)
Links to Code Toggle
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub Toggle
DagsHub (What is DagsHub?)
GotitPub Toggle
Gotit.pub (What is GotitPub?)
Huggingface Toggle
Hugging Face (What is Huggingface?)
ScienceCast Toggle
ScienceCast (What is ScienceCast?)
Demos
Demos
Replicate Toggle
Replicate (What is Replicate?)
Spaces Toggle
Hugging Face Spaces (What is Spaces?)
Spaces Toggle
TXYZ.AI (What is TXYZ.AI?)
Related Papers
Recommenders and Search Tools
Link to Influence Flower
Influence Flower (What are Influence Flowers?)
Core recommender toggle
CORE Recommender (What is CORE?)
Author
Venue
Institution
Topic
About arXivLabs
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Which authors of this paper are endorsers? |
Disable MathJax (What is MathJax?)