We built HNSW directly on the embedding lists, with no encoding step in between. 𝗧𝗵𝗲 ... - 惯性聚合

推荐订阅源

Forbes - Security

大猫的无限游戏

让小产品的独立变现更简单 - ezindie.com

美团技术团队

博客园 - 聂微东

博客园 - 叶小钗

Security Latest

Threat Intelligence Blog | Flashpoint

博客园_首页

Java Code Geeks

宝玉的分享

Cyber Attacks, Cyber Crime and Cyber Security

Privacy International News Feed

CXSECURITY Database RSS Feed - CXSecurity.com

Threat Research - Cisco Blogs

The Hacker News

LINUX DO - 热门话题

freeCodeCamp Programming Tutorials: Python, JavaScript, Git & More

Darknet – Hacking Tools, Hacker News & Cyber Security

Last Week in AI

人人都是产品经理

酷壳 – CoolShell

Tor Project blog

罗磊的独立博客

Vulnerabilities – Threatpost

Apple Machine Learning Research

钛媒体：引领未来商业与生活新知

有赞技术团队

Hugging Face - Blog

奇客Solidot–传递最新科技情报

The Exploit Database - CXSecurity.com

博客园 - 【当耐特】

Schneier on Security

Secure Thoughts

cs.CL updates on arXiv.org

博客园 - 司徒正美

Milvus(@milvusio)

Milvus (@milvusio) on X 𝗠𝗶𝗹𝘃𝘂𝘀 𝗰𝗼𝗺𝗽𝗮𝗰𝘁𝘀 𝘀𝗲𝗴𝗺𝗲𝗻𝘁𝘀 𝗱𝘂𝗿𝗶𝗻𝗴 𝗻𝗼𝗿𝗺𝗮𝗹 𝗼𝗽𝗲𝗿𝗮𝘁𝗶𝗼𝗻𝘀, ... 𝗡𝗲𝗮𝗿𝗹𝘆 𝗵𝗮𝗹𝗳 𝘆𝗼𝘂𝗿 𝗠𝗶𝗹𝘃𝘂𝘀 𝘀𝗲𝗮𝗿𝗰𝗵 𝗰𝗮𝗽𝗮𝗰𝗶𝘁𝘆 𝗺𝗮𝘆 𝗯𝗲 ... 𝗔𝗱𝗱𝗶𝗻𝗴 𝗮𝗻 𝗔𝗜 𝗮𝗴𝗲𝗻𝘁 𝘁𝗼 𝗱𝗮𝘁𝗮𝗯𝗮𝘀𝗲 𝘁𝗼𝗼𝗹𝗶𝗻𝗴 𝗿𝗮𝗶𝘀𝗲𝘀 𝗮𝗻 ... 𝗠𝗶𝗹𝘃𝘂𝘀 𝗷𝘂𝘀𝘁 𝗰𝗿𝗼𝘀𝘀𝗲𝗱 𝟰𝟱,𝟬𝟬𝟬 𝗚𝗶𝘁𝗛𝘂𝗯 𝘀𝘁𝗮𝗿𝘀 ⭐️ A huge thank you ❤️ to ... MFS 01: https://t.co/PVPqlHy7CR MFS 03: https://t.co/ksYX81kOHA 𝗪𝗲 𝗼𝗽𝗲𝗻-𝘀𝗼𝘂𝗿𝗰𝗲𝗱 𝗠𝗙𝗦 (𝗠𝘂𝗹𝘁𝗶-𝘀𝗼𝘂𝗿𝗰𝗲 𝗙𝗶𝗹𝗲-𝗹𝗶𝗸𝗲 𝗦𝗲𝗮𝗿𝗰𝗵), a tool... ⏰ Coming up very soon: 𝗙𝗿𝗼𝗺 𝗩𝗲𝗰𝘁𝗼𝗿 𝗗𝗮𝘁𝗮𝗯𝗮𝘀𝗲 𝘁𝗼 𝗩𝗲𝗰𝘁𝗼𝗿 𝗟𝗮𝗸𝗲𝗯𝗮𝘀𝗲 sta... 𝗪𝗲 𝗼𝗽𝗲𝗻-𝘀𝗼𝘂𝗿𝗰𝗲𝗱 𝗢𝗽𝗲𝗻 𝗧𝗮𝗴, 𝗮𝗻 𝗼𝗽𝗲𝗻 𝘃𝗲𝗿𝘀𝗶𝗼𝗻 𝗼𝗳 𝗖𝗹𝗮𝘂𝗱𝗲 𝗧𝗮𝗴.... We kept seeing the same pattern with agents: when the same agent lived inside IM instead of a termin... ⏰ Quick reminder: our webinar, 𝗙𝗿𝗼𝗺 𝗩𝗲𝗰𝘁𝗼𝗿 𝗗𝗮𝘁𝗮𝗯𝗮𝘀𝗲 𝘁𝗼 𝗩𝗲𝗰𝘁𝗼𝗿 ... 𝗚𝗿𝗮𝗻𝘁𝗶𝗻𝗴 𝗮 𝘁𝗲𝗮𝗺𝗺𝗮𝘁𝗲 𝗿𝗲𝗮𝗱-𝗼𝗻𝗹𝘆 𝗮𝗰𝗰𝗲𝘀𝘀 𝘁𝗼 𝗼𝗻𝗲 𝗠𝗶𝗹𝘃𝘂𝘀 ... 𝗔 𝘀𝗶𝗻𝗴𝗹𝗲 𝟭–𝟱 𝘀𝗰𝗼𝗿𝗲 𝗶𝘀 𝗮 𝗯𝗮𝗱 𝘄𝗮𝘆 𝘁𝗼 𝗷𝘂𝗱𝗴𝗲 𝗥𝗔𝗚 𝗾𝘂𝗮𝗹𝗶𝘁𝘆 𝗶𝗻 ... 𝗟𝗟𝗠𝘀 𝗸𝗲𝗲𝗽 𝗰𝗶𝘁𝗶𝗻𝗴 𝘀𝗼𝘂𝗿𝗰𝗲𝘀 𝘁𝗵𝗮𝘁 𝗱𝗼𝗻'𝘁 𝘀𝗮𝘆 𝘄𝗵𝗮𝘁 𝘁𝗵𝗲𝘆 ... Attu used to feel like a lightweight console for one Milvus instance. 𝗔𝘁𝘁𝘂 𝟯.𝟬 𝗕𝗲𝘁𝗮 𝗶𝘀 ... 𝗔 𝗹𝗼𝘁 𝗼𝗳 𝗱𝗮𝘁𝗮𝗯𝗮𝘀𝗲 𝗔𝗜 𝗳𝗲𝗮𝘁𝘂𝗿𝗲𝘀 𝘀𝘁𝗼𝗽 𝗮𝘁 𝗮𝗻𝘀𝘄𝗲𝗿𝗶𝗻𝗴 ... 𝗬𝗼𝘂𝗿 𝘁𝗲𝘀𝘁 𝘀𝗲𝘁 𝘀𝗵𝗼𝘄𝘀 𝘀𝘁𝗿𝗼𝗻𝗴 𝗿𝗲𝗰𝗮𝗹𝗹, 𝗯𝘂𝘁 𝗿𝗲𝘁𝗿𝗶𝗲𝘃𝗮𝗹 𝗺𝗮𝘆 ... In text retrieval, multi-vector's advantage is conditional. On short documents and simple queries, a... 𝗜𝗳 𝘆𝗼𝘂 𝗿𝘂𝗻 𝗺𝗼𝗿𝗲 𝘁𝗵𝗮𝗻 𝗼𝗻𝗲 𝗠𝗶𝗹𝘃𝘂𝘀 𝗰𝗹𝘂𝘀𝘁𝗲𝗿, 𝘆𝗼𝘂 𝗸𝗻𝗼𝘄 𝘁𝗵𝗲 ... A user tells your support chatbot “my service keeps freezing.” Your knowledge base has the right art... 𝗠𝗶𝗹𝘃𝘂𝘀 𝗵𝗮𝘀 𝗿𝗲𝗹𝗲𝗮𝘀𝗲𝗱 𝗔𝘁𝘁𝘂 𝟯.𝟬 𝗕𝗲𝘁𝗮, 𝗮 𝗿𝗲𝗯𝘂𝗶𝗹𝘁 ... 𝗗𝗼 𝘆𝗼𝘂 𝗸𝗻𝗼𝘄 𝗵𝗼𝘄 𝗺𝗮𝗻𝘆 𝗯𝗶𝗿𝗱𝘀 𝘁𝗵𝗲𝗿𝗲 𝗮𝗿𝗲 𝗶𝗻 𝘁𝗵𝗲 𝗠𝗶𝗹𝘃𝘂𝘀 𝗮𝗻𝗱 ... 𝗟𝗘𝗠𝗨𝗥 𝗶𝘀 𝘁𝗵𝗲 𝗺𝘂𝗹𝘁𝗶-𝘃𝗲𝗰𝘁𝗼𝗿 𝘀𝘁𝗿𝗮𝘁𝗲𝗴𝘆 𝘁𝗵𝗮𝘁 𝗹𝗼𝗼𝗸𝘀 𝗲𝗮𝘀𝗶𝗲𝘀𝘁 ... 𝗛𝗼𝘄 𝗱𝗼𝗲𝘀 𝗮 𝘀𝘁𝗼𝗰𝗸 𝗺𝗲𝗱𝗶𝗮 𝗽𝗹𝗮𝘁𝗳𝗼𝗿𝗺 𝘄𝗶𝘁𝗵 𝟮𝟬𝟬𝗠+ 𝘃𝗶𝘀𝘂𝗮𝗹 ... 𝗪𝗲 𝗰𝗹𝗮𝘀𝘀𝗶𝗳𝘆 𝗛𝗲𝗿𝗺𝗲𝘀 𝘄𝗶𝘁𝗵 𝘀𝗸𝗶𝗹𝗹𝘀 𝗮𝘀 𝗮𝗻 𝗟𝟯. 𝗛𝗲𝗿𝗲'𝘀 𝘄𝗵𝘆, 𝗮𝗻𝗱 ... 📣 𝗢𝘂𝗿 𝘄𝗲𝗯𝗶𝗻𝗮𝗿 𝗮𝗯𝗼𝘂𝘁 𝗠𝗶𝗹𝘃𝘂𝘀 𝟯.𝟬 𝗶𝘀 𝗵𝗮𝗽𝗽𝗲𝗻𝗶𝗻𝗴 𝘁𝗼𝗱𝗮𝘆! Join us... 𝗬𝗼𝘂 𝗰𝗮𝗻 𝗿𝘂𝗻 𝟮𝟱 𝗺𝗶𝗹𝗹𝗶𝗼𝗻 𝘃𝗲𝗰𝘁𝗼𝗿𝘀 𝗶𝗻 𝗠𝗶𝗹𝘃𝘂𝘀 𝘂𝘀𝗶𝗻𝗴 𝘂𝗻𝗱𝗲𝗿 ... 𝗪𝗲𝗯𝗶𝗻𝗮𝗿 𝗥𝗲𝗺𝗶𝗻𝗱𝗲𝗿 | 𝗪𝗵𝗮𝘁’𝘀 𝗡𝗲𝘄 𝗶𝗻 𝗠𝗶𝗹𝘃𝘂𝘀 𝟯.𝟬: 𝗟𝗶𝘃𝗲 ... With the same multi-vector model, and the same dataset, nDCG@10 can drop from 0.701 to 0.109 — rough... 𝗧𝗵𝗲𝗿𝗲 𝗮𝗿𝗲 𝘁𝗵𝗿𝗲𝗲 𝗰𝗼𝗺𝗺𝗼𝗻 𝘄𝗮𝘆𝘀 𝘁𝗼 𝗰𝗵𝘂𝗻𝗸 𝗱𝗼𝗰𝘂𝗺𝗲𝗻𝘁𝘀 𝗳𝗼𝗿 𝗥𝗔𝗚.... Your RAG tested well and went live, but recall is getting worse. 𝗧𝗵𝗿𝗲𝗲 𝗰𝗼𝗺𝗺𝗼𝗻 ... Vector search works well when semantic meaning matters, but it can be unreliable with exact terms. S... 🌐 𝗪𝗲𝗯𝗶𝗻𝗮𝗿 | 𝗪𝗵𝗮𝘁’𝘀 𝗡𝗲𝘄 𝗶𝗻 𝗠𝗶𝗹𝘃𝘂𝘀 𝟯.𝟬: 𝗟𝗶𝘃𝗲 𝗪𝗮𝗹𝗸𝘁𝗵𝗿𝗼𝘂𝗴𝗵 & ... 𝗖𝗹𝗮𝘂𝗱𝗲 𝗢𝗽𝘂𝘀 𝟰.𝟴 brings longer independent work, sharper judgment, and stronger self-chec... When RAG gives a wrong answer, the team's first move is usually a bigger model or more prompt tweaki... 𝗜𝗻 𝗮 𝗹𝗼𝗻𝗴-𝗿𝘂𝗻𝗻𝗶𝗻𝗴 𝗥𝗔𝗚 𝘀𝘆𝘀𝘁𝗲𝗺, 𝘁𝗵𝗲 𝗺𝗼𝘀𝘁 𝗱𝗮𝗻𝗴𝗲𝗿𝗼𝘂𝘀 𝗯𝘂𝗴 ... 𝗖𝗥𝗔𝗚 𝗰𝗮𝗻 𝗺𝗮𝗸𝗲 𝘆𝗼𝘂𝗿 𝗥𝗔𝗚 𝘀𝘂𝗿𝗳𝗮𝗰𝗲 𝘁𝗵𝗲 𝗮𝗻𝘀𝘄𝗲𝗿 𝘆𝗼𝘂 𝗮𝗰𝘁𝘂𝗮𝗹𝗹𝘆 ... 𝗧𝗵𝗲𝗿𝗲'𝘀 𝗽𝗿𝗼𝗯𝗮𝗯𝗹𝘆 𝗮 𝗯𝗹𝗶𝗻𝗱 𝘀𝗽𝗼𝘁 𝗶𝗻 𝘆𝗼𝘂𝗿 𝗥𝗔𝗚 𝗽𝗶𝗽𝗲𝗹𝗶𝗻𝗲: 𝗶𝘁 ... This time, 𝗤𝘄𝗲𝗻𝟯.𝟳-𝗠𝗮𝘅 was not released with open weights. But for enterprise agents, it is... Korean memory stocks are going crazy. SK Hynix has nearly tripled since the end of 2025. If you run ... ❓ 𝗛𝗼𝘄 𝗱𝗼 𝘆𝗼𝘂 𝗸𝗲𝗲𝗽 𝗳𝗶𝗹𝘁𝗲𝗿𝗲𝗱 𝘃𝗲𝗰𝘁𝗼𝗿 𝘀𝗲𝗮𝗿𝗰𝗵 𝗳𝗮𝘀𝘁 𝗮𝗻𝗱 ... A lot of the "RAG is dead" arguments have some truth: traditional RAG is a poor fit for agentic work... 𝗙𝗼𝗿 𝗮 𝘄𝗵𝗶𝗹𝗲, 𝘄𝗲 𝗸𝗲𝗽𝘁 𝗵𝗲𝗮𝗿𝗶𝗻𝗴 𝘁𝗵𝗲 𝘀𝗮𝗺𝗲 𝗰𝗼𝗺𝗽𝗹𝗮𝗶𝗻𝘁 𝗳𝗿𝗼𝗺 𝗲-... ❓ 𝗛𝗼𝘄 𝗱𝗼 𝘆𝗼𝘂 𝗿𝗲𝗱𝘂𝗰𝗲 𝘀𝗲𝗿𝘃𝗶𝗻𝗴 𝗰𝗼𝘀𝘁𝘀 𝘄𝗶𝘁𝗵𝗼𝘂𝘁 𝗺𝗮𝗸𝗶𝗻𝗴 ... Most AI teams do not start with a blank slate. They already have data in object storage, pipelines, ... If you’ve used Milvus, you probably know Attu. It’s the UI many developers open when they want to i... At last month’s Unstructured Data Meetup London, Jiang Chen, our Head of Developer Relations, broke ... In this previous post (https://t.co/77L5mrn5q7), we talked about why multi-vector models sometimes h... When search results aren't ideal, the default instinct is to filter them out. 𝗜𝗻 ...

We built HNSW directly on the embedding lists, with no encoding step in between. 𝗧𝗵𝗲 ...

Milvus (@mil · 2026-06-23 · via Milvus(@milvusio)

此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。原文来自 — 版权归原作者所有。