揭秘AI Agent网络的三层容错机制，从指数退避到熔断再到管线重规划，手把手教你构建自愈系统。
In a single-agent system, failure is simple: the agent errors, you retry. In multi-agent systems, failure is a graph problem. The Cascade Failure Prob…
腾讯如何用SiriusDeliver实现数据仓库交付自动化，一窥工业级数据工程的最新实践。
arXiv:2608.09185v1 Announce Type: cross Abstract: Enterprise data warehouses (DWs) support business-critical analytics, but warehouse task delivery re…
重磅预印本！终端智能体或将成为企业自动化的最优解，51页详实论证，AI Agent落地方向必读。
arXiv:2604.00073v3 Announce Type: replace-cross Abstract: There has been growing interest in building agents that can interact with digital platforms …
行业首个智算系统架构联盟成立，发布《超节点系统架构规范》，多方协同打造自主可控算力底座。
LLM时代维基维护面临知识漂移与内容矛盾，本文系统梳理了审核机制与系统架构的挑战与解法
Article URL: https://www.glukhov.org/knowledge-management/knowledge-systems-architectures/compiled-knowledge/llm-wiki-maintenance-knowledge-drift/ Com…
这篇arXiv论文提出面向Agentic LLM系统的共享选择性持久记忆机制，解决多智能体协作中的记忆共享与选择性持久化问题，是AI大模型研究前沿。
arXiv:2607.09493v1 Announce Type: new Abstract: Agentic LLM systems that generate code through multi-turn tool use face a fundamental context problem:…
复合AI系统聚合能力的边界在哪？这篇论文给出理论答案，值得关注AI架构的开发者细读。
arXiv:2602.21556v2 Announce Type: replace Abstract: When designing compound AI systems, a common approach is to query multiple copies of the same mode…
从基因调控网络到智能体软件，揭示生物启发的控制架构新思路
arXiv:2607.04240v1 Announce Type: new Abstract: The transition of Large Language Models (LLMs) from passive generators to autonomous agents has introd…
从单脑到五脑，揭秘 Halo 拆分多智能体的实战架构，看侦察、利用与自校正如何分而治之。
For most of Halo's life, it's been a single agent doing everything. One loop, one model, one context window juggling reconnaissance, exploitation, and…
深入剖析Claude Code的设计哲学，系统探讨当前与未来AI Agent架构的设计空间，一篇不容错过的技术佳作。
arXiv:2604.14228v2 Announce Type: replace-cross Abstract: Claude Code is an agentic coding tool that can run shell commands, edit files, and call exte…
揭秘Agentic AI背后“线路工程”的精髓，如何构建包裹LLM的骨架以应对现实不完美
Harness Engineering 101 — สิ่งที่อยู่ใต้พรมของ Agentic AI บทความก่อนเราคุยกันเรื่อง "จาก LLM เปล่า → Agentic AI" แบบ 7 layer คราวนี้มาดูว่าภายในแต่ละ …
用真实案例讲透企业级AI交易对账系统构建，含架构思路与落地经验，值得技术决策者一读。
From Unstructured Bank Statements to Automated SAP Reconciliation For years, I've read articles claiming that AI would revolutionize enterprise financ…
多智能体LLM如何高效协调？本文提出受治理的共享内存机制，为解决Agent间信息共享与一致性难题提供新方案。
arXiv:2606.24535v1 Announce Type: new Abstract: Multi-agent LLM environments require robust mechanisms for shared knowledge management. This paper for…
图数据库驱动的健康AI系统，在多学科肿瘤诊疗中落地，架构清晰值得借鉴。
arXiv:2606.22692v1 Announce Type: new Abstract: We introduce VISTA Architect, a database-oriented AI architecture for integrating large language model…
DeepMind提出“多层安全”新理念，将防护嵌入AI系统每个层级，而非简单打补丁。
Google DeepMind, yapay zeka ajanlarının birbirleriyle konuştuğu çoklu ajan sistemleri (multi-agent systems) tüm dünyaya yayılmadan önce güvenliği sağl…
实测对比Agent记忆的推拉模式，开源项目帮你搞清楚哪个更适合你的AI系统。
Article URL: https://github.com/H-XX-D/recall-memory-substrate Comments URL: https://news.ycombinator.com/item?id=48580203 Points: 1 # Comments: 0
颠覆认知：AI“幻觉”的根源不在模型，而在糟糕的系统架构设计。
Everyone is talking about hallucination. That's the wrong diagnosis. Hallucination isn't a bug. It's the mechanism. Turn the temperature down far enou…
Anthropic的Mythos模型训练数据源自哪里？社区讨论触及RCA训练数据公开的难题。
This question came up from a tangential discussion. Some one was asking suggestions about systems architecture. I suggested them to look into RCA (roo…
深入解析LLM控制平面概念，区分数据平面与控制平面，助你构建生产级AI系统。
Article URL: https://blog.mozilla.ai/what-is-an-llm-control-plane/ Comments URL: https://news.ycombinator.com/item?id=48508913 Points: 1 # Comments: 0