BriefGPT - AI 论文速递 BriefGPT - AI 论文速递 ·

GLADMamba: Unsupervised Graph-Level Anomaly Detection Based on Selective State Space Model

💡 原文英文，约100词，阅读约需1分钟。
📝

内容提要

本研究提出了GLADMamba框架，解决了无监督图级异常检测中的长距离依赖和光谱信息忽视问题。通过选择性状态空间模型，GLADMamba在12个真实数据集上显著提升了检测性能。

🏷️

标签

GLADMamba graph model 图级异常检测 无监督 选择性状态空间模型 长距离依赖
➡️

继续阅读

  1. State of Consumer Africa 2026: Meeting the pragmatic shopper
    Household budgets in Africa are tight and channels increasingly fragmented. F...
  2. 当关系本身构成证据时，请使用Graph RAG
    向量检索擅长按语义找相似文本，但无法证明事实关联。当答案依赖多跳关系（如漏洞库→服务→客户→合同→通知政策）时，应引入Graph RAG：将业务记录建模为...
  3. 为Graph-RAG添加时间推理：追踪事实的新鲜度与陈旧度
    文章介绍为Graph-RAG加入轻量级时间推理层：将传统主谓宾三元组扩展为带时间戳的四元组并存入时间图，用指数衰减计算事实新鲜度权重，按查询日期对冲突事实...
  4. TypeSafe AI Releases Jev: A Decision-Only Model That Returns Typed Probabilities Instead of Text
    TypeSafe AI, founded by former OpenAI researcher Diogo Almeida, has introduce...
  5. Engineering Production Systems for an Agentic Era: QCon San Francisco 2026
    QCon San Francisco 2026 will bring together practitioners from Airbnb, OpenAI...
  6. Article: From Reusable to Regeneratable: Rethinking the Shared UI Component Library
    Regeneration has changed reuse economics. A model pointed at a solid design s...
👤 个人中心
在公众号发送验证码完成验证