让多智能体系统在冲突中做出可信决策，避免盲目投票。
LatticeMind: A Conflict-Aware Memory Primitive for Multi-Agent Systems
- 用结构化记忆实时处理矛盾，写入时即判断冲突状态。
- 在无源提示的测试中准确率达97%，远超基线61%。
- 适合需要持续信任推理链的复杂协作任务。
多智能体大模型系统常因缺乏持久机制来决定哪些相互矛盾的主张应被信任而失败。多数投票、辩论和裁判选择均不记录哪个主张胜出、是否存疑，或为何后续更新取代前者。我们提出LatticeMind，一种冲突感知的结构化记忆，可在写入时处理矛盾。它显式维护条目状态，执行低成本符号冲突检查，并仅对未解决的语义冲突调用LLM进行协调。在移除来源名称提示的标签盲冲突评估中，LatticeMind达到0.97准确率，远超最强聚合基线的0.61，差异在配对McNemar检验下显著（p<10⁻⁶）。消融实验显示，移除检查器或协调器会导致性能下降12至14点。在四个次要规划基准上表现参差：在三个任务上优于简单合并，在需迭代搜索的任务中仍不如深度推理方法。
原文摘要 · Abstract (English)
Multi-agent LLM systems often fail not for lack of candidate answers, but because they have no persistent mechanism for deciding which incompatible claim should currently be trusted. Majority vote, debate, and judge-based selection choose an output without recording which claim wins, which is contested, or why a later update supersedes it. We present \term{LatticeMind}, a conflict-aware structured memory that handles contradiction at write time. It maintains explicit item status, applies cheap symbolic conflict checks, and invokes LLM reconciliation only for unresolved semantic cases. On a label-blind ConflictBank evaluation that removes source-name hints, LatticeMind reaches 0.97 accuracy versus 0.61 for the strongest aggregation baseline, with the gap significant at $p<10^{-6}$ by paired McNemar test. Ablations show that removing the checker or the reconciler costs 12 to 14 points. On four secondary planning benchmarks the picture is mixed: LatticeMind beats naive merge on three of four, but does not replace deliberation methods on tasks rewarding iterative search.
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。