提出多智能体知识库的协商式治理协议，提升系统在压力下的稳定性。
Deliberative Curation: A Protocol for Multi-Agent Knowledge Bases
- 用状态转移系统管理知识生命周期，结合声誉投票与分层惩罚机制。
- 在压力下精度优于传统多数投票，退化速度慢三倍以上。
- 适合构建抗干扰、自组织的知识协作系统，尤其适用于开放环境。
随着人工智能代理从孤立工具转向共享知识生态中的协作参与者，集体知识治理成为关键挑战。人类平台治理机制无法直接迁移：代理的无状态性削弱威慑性制裁，模型同质性违背群体智慧的独立性假设，谄媚行为破坏协商共识。本文提出一种包含三层的协商式治理协议：（1）将知识资产生命周期形式化为带标签的转移系统；（2）融合贝塔声誉与特征信任放大效应的加权协商投票；（3）针对无状态代理设计的渐进式制裁机制，区分故障与恶意行为。通过100个代理的仿真评估，涵盖七种行为原型，在两种压力场景下进行30组种子测试（配对t检验）。该协议在良性条件下略低精度，但在压力下显著更稳健：中等压力下准确率0.826对比多数投票的0.791（p<0.001），极端压力下达0.807比0.740（p<0.001），退化速度仅为多数投票的约三分之一。消融分析表明，提交-揭示投票隐匿是单个最有效组件，带来8.2–8.6个百分点精度提升（p<0.001），超过声誉加权与协商机制之和。渐进式制裁未在模拟中触发，尚待实证验证。
原文摘要 · Abstract (English)
As AI agents transition from isolated tools to collaborative participants in shared knowledge ecosystems, governing collective knowledge curation becomes a critical challenge. Human platform governance mechanisms do not transfer directly: agent statelessness undermines deterrence-based sanctions, model homogeneity violates independence assumptions underlying crowd wisdom, and sycophancy collapses deliberative consensus. We propose a deliberative curation protocol combining three governance layers: (1) a knowledge artifact lifecycle formalized as a labeled transition system; (2) reputation-weighted deliberative voting integrating Beta Reputation with EigenTrust amplification; and (3) graduated sanctions adapted for stateless agents, including broken agent handling distinguishing malfunction from adversarial behavior. We evaluate the protocol through agent-based simulation with 100 agents across seven behavioral archetypes under two adversity scenarios (30 seeds, paired t-tests). The protocol trades modest precision under benign conditions for substantially better resilience under adversity: 0.826 vs 0.791 for majority vote under moderate adversity (p<0.001), widening to 0.807 vs 0.740 under stress (p<0.001). The protocol degrades roughly three times more slowly than majority vote. Ablation analysis identifies commit-reveal vote concealment as the most impactful single component (8.2-8.6pp precision improvement, p<0.001), outperforming reputation weighting and deliberation combined. Graduated sanctions were not exercised in simulation and remain empirically unvalidated.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。