Σ-Mem: An Online Reliability Memory for LLM-based Multi-Agent Systems

  • 类型：arxiv
  • 标识：2607.27958
  • 链接：https://arxiv.org/abs/2607.27958
  • 主分类：agent
  • 形态：method
  • 被引：0
  • 被引来源：Semantic Scholar + OpenAlex
  • S2被引：0
  • OpenAlex被引：0
  • 影响力被引：0
  • TLDR：Memory is central to long-horizon LLM agents, yet existing memory systems primarily preserve interaction content rather than modeling which agents can be trusted and under what conditions. This limitation is particularly important in multi-agent systems, where a central model may be unable to directly verify plausible or correlated peer responses. We introduce Σ-Mem, an online reliability memory that records historical competence evidence for individual peers and peer relationship evidence across the peer set. Both forms of evidence are maintained as real symmetric states and updated from post
  • OpenAlex ID：W7171979022
  • OpenAlex DOI：10.48550/arxiv.2607.27958
  • DOI：10.48550/arxiv.2607.27958
  • DOI来源：OpenAlex
  • 开放获取：green
  • 开放获取链接：https://doi.org/10.48550/arxiv.2607.27958
  • OpenAlex更新：2026-08-26
  • 待LLM分类：否
  • 标题中文：Σ-Mem：基于 LLM 的多智能体系统的在线可靠性记忆
  • TLDR中文：记忆是长时程 LLM 智能体的核心，但现有记忆系统主要保存交互内容，而未建模哪些智能体在何种条件下可信。这一局限在多智能体系统中尤为关键，因为中心模型可能无法直接验证来自对等方、看似合理或相关的响应。我们提出 Σ-Mem，一种在线可靠性记忆，记录单个对等方的历史能力证据以及跨对等集的对等关系证据。两种证据均以实对称状态形式维护，并基于后
  • 来源文件：
  • /inbox/tom/_candidates/2026-07-31-agent-rag-longcontext-candidates.json
  • /inbox/tom/_candidates/2026-08-01-agent-rag-longcontext-candidates.json
  • /inbox/tom/_candidates/2026-08-02-agent-rag-longcontext-candidates.json
  • /inbox/tom/_candidates/2026-08-03-agent-rag-longcontext-candidates.json
  • /inbox/tom/_candidates/2026-08-03-agent-memory-tool-use-candidates.json
  • [S2 enrich]
  • [OpenAlex backfill]