<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0"><channel><title>AI工具与模型HOT</title><link>https://staroltest.github.io/ai-tool-model-hot/</link><description>每天从 21 个信源收 AI 工具与模型动态，按规则打分精选，只显示标题、一句简介与原文链接，不转载全文。</description><language>zh-CN</language><lastBuildDate>Tue, 17 Nov 2026 11:11:08 GMT</lastBuildDate><item><title>NVIDIA Vera Rubin 平台 MLPerf 首秀，训练与推理全面领先</title><link>https://blogs.nvidia.com/blog/mlperf-training-endpoints-vera-rubin</link><description>Vera Rubin 平台首次参加 MLPerf 就在训练、后训练和智能体推理上拿下领先成绩，英伟达包揽了全部项目。</description><pubDate>Mon, 16 Nov 2026 16:00:25 GMT</pubDate><guid isPermaLink="false">0f2a855bc6a8fcf25d6065605b52cc9d13fe5cc6</guid><source url="https://blogs.nvidia.com/blog/mlperf-training-endpoints-vera-rubin">NVIDIA Blog</source></item><item><title>前沿科学：1800 多位科学家怎么用 ChatGPT 做研究</title><link>https://openai.com/index/science-at-the-frontier</link><description>OpenAI 调查了 1800 多位科学家的日常用法，展示 AI 如何改变研究流程、效率和对发现的期待。</description><pubDate>Mon, 16 Nov 2026 12:00:00 GMT</pubDate><guid isPermaLink="false">f891bd93f3d21469060271c5a85b2a7a2e8b4138</guid><source url="https://openai.com/index/science-at-the-frontier">OpenAI News</source></item><item><title>NVIDIA CUDA-X 扩展库，让智能体加速科学发现</title><link>https://blogs.nvidia.com/blog/cuda-x-libraries-agents-accelerate-scientific-discovery</link><description>新扩的 CUDA-X 库帮研究者在天体物理、材料科学等领域更快地筛海量数据，智能体也能直接调用这些工具。</description><pubDate>Mon, 16 Nov 2026 16:00:16 GMT</pubDate><guid isPermaLink="false">8192f1094925b8739815e3f9d9444b2ff594ba2f</guid><source url="https://blogs.nvidia.com/blog/cuda-x-libraries-agents-accelerate-scientific-discovery">NVIDIA Blog</source></item><item><title>NVIDIA 发布 Rubin HPC GPU，为科学、工程与分析加速智能体</title><link>https://blogs.nvidia.com/blog/rubin-hpc</link><description>这款 Vera Rubin 平台的新 GPU 面向仿真和数据处理，意在让智能体在科研和工程流程里跑得更快。</description><pubDate>Mon, 16 Nov 2026 16:00:01 GMT</pubDate><guid isPermaLink="false">6cab091c059af43fb71f81530930f3b85777319c</guid><source url="https://blogs.nvidia.com/blog/rubin-hpc">NVIDIA Blog</source></item><item><title>LangChain 完成 3 亿美元 C 轮融资，估值 30 亿美元</title><link>https://www.langchain.com/blog/series-c</link><description>LangChain 拿到 3 亿美元新融资，用来帮企业用 LangSmith 掌握自己的智能体能力。</description><pubDate>Mon, 16 Nov 2026 23:53:29 GMT</pubDate><guid isPermaLink="false">1742558d91d037f5dac22c418ea1ee9137d71f48</guid><source url="https://www.langchain.com/blog/series-c">LangChain Blog</source></item><item><title>OpenAI 扩大面向科学家的 AI 使用权限</title><link>https://openai.com/index/expanding-chatgpt-for-academic-researchers</link><description>OpenAI 扩展了 ChatGPT 学术研究者计划，将向全球 20 万名研究人员免费开放其前沿模型。</description><pubDate>Mon, 16 Nov 2026 17:00:00 GMT</pubDate><guid isPermaLink="false">f26f2a4dfd6b9b957c1544d02f570339b8f668d2</guid><source url="https://openai.com/index/expanding-chatgpt-for-academic-researchers">OpenAI News</source></item><item><title>OpenAI 推出新工具与支持，加强打击网络儿童性剥削</title><link>https://openai.com/index/strengthening-online-child-safety</link><description>OpenAI 推出新工具、实用指南与合作项目，帮助预防、发现和应对网络儿童性剥削与性虐待。</description><pubDate>Mon, 16 Nov 2026 17:00:00 GMT</pubDate><guid isPermaLink="false">b1c71d07875ac772d826f89afdf2886a79e196cc</guid><source url="https://openai.com/index/strengthening-online-child-safety">OpenAI News</source></item><item><title>英伟达扩展 Slurm，把量子计算带入超算中心</title><link>https://blogs.nvidia.com/blog/slurm-quantum-computing-supercomputing</link><description>随着超算中心部署量子处理器的数量五年增长近七倍，英伟达把 Slurm 扩展到量子计算领域，让量子与经典超算能协同调度。</description><pubDate>Mon, 16 Nov 2026 16:00:47 GMT</pubDate><guid isPermaLink="false">0be2ab6226e78541de53fadb4bedd8b18e0f461f</guid><source url="https://blogs.nvidia.com/blog/slurm-quantum-computing-supercomputing">NVIDIA Blog</source></item><item><title>Bedrock 新增 Agentic Retrieve，让 AI 智能体带上长期记忆</title><link>https://aws.amazon.com/blogs/machine-learning/build-memory-aware-ai-agents-with-agentic-retrieve-on-amazon-bedrock-managed-knowledge-base</link><description>Bedrock 把对话记录和提炼出的事实统一存进知识库，智能体一次调用就能带上长期记忆来回答。</description><pubDate>Mon, 16 Nov 2026 22:43:30 GMT</pubDate><guid isPermaLink="false">d218f9a95b85493a3514327d4ce07ea6c3e70636</guid><source url="https://aws.amazon.com/blogs/machine-learning/build-memory-aware-ai-agents-with-agentic-retrieve-on-amazon-bedrock-managed-knowledge-base">AWS Machine Learning Blog</source></item><item><title>Duco 用 Bedrock AgentCore 重做金融数据自动化</title><link>https://aws.amazon.com/blogs/machine-learning/how-duco-transforms-financial-data-automation-with-amazon-bedrock-agentcore</link><description>这家服务头部银行的数据平台把智能体栈搬上 Bedrock，用多模型策略和严格的租户隔离来处理金融数据。</description><pubDate>Mon, 16 Nov 2026 22:06:08 GMT</pubDate><guid isPermaLink="false">94ec9ff9d4729a9eb2f9bd35f0316e46c9857d61</guid><source url="https://aws.amazon.com/blogs/machine-learning/how-duco-transforms-financial-data-automation-with-amazon-bedrock-agentcore">AWS Machine Learning Blog</source></item><item><title>多智能体设计模式与反模式分类：来自制造业的经验</title><link>https://aws.amazon.com/blogs/machine-learning/taxonomy-of-multi-agent-design-patterns-anti-patterns-and-lessons-from-manufacturing</link><description>文章把多智能体协作当成分布式系统问题来拆解，给出一套实战验证过的模式、反模式和取舍框架。</description><pubDate>Mon, 16 Nov 2026 21:53:46 GMT</pubDate><guid isPermaLink="false">0761da68e54fd0db019f399e814ef824491e43e9</guid><source url="https://aws.amazon.com/blogs/machine-learning/taxonomy-of-multi-agent-design-patterns-anti-patterns-and-lessons-from-manufacturing">AWS Machine Learning Blog</source></item><item><title>MoEngage 用 Amazon Bedrock AgentCore 搭建智能体客户运营</title><link>https://aws.amazon.com/blogs/machine-learning/how-moengage-built-agentic-customer-engagement-on-amazon-bedrock-agentcore</link><description>MoEngage 把营销和产品团队的活动起草、分析和调整都收进一次对话里，用 Bedrock 上的智能体来跑客户运营。</description><pubDate>Mon, 16 Nov 2026 18:16:07 GMT</pubDate><guid isPermaLink="false">9567fc17c1fdcddfad2f51d66f8650065fd212ba</guid><source url="https://aws.amazon.com/blogs/machine-learning/how-moengage-built-agentic-customer-engagement-on-amazon-bedrock-agentcore">AWS Machine Learning Blog</source></item><item><title>Amazon Bedrock 托管知识库新增表格搜索，可直接从表格中找答案</title><link>https://aws.amazon.com/blogs/machine-learning/get-answers-from-tables-with-tabular-search-in-amazon-bedrock-managed-knowledge-base</link><description>Bedrock 托管知识库现在支持表格搜索，智能体检索能直接查 CSV、XLSX 与 XLS 文件里的数值并做计算。</description><pubDate>Mon, 16 Nov 2026 16:12:16 GMT</pubDate><guid isPermaLink="false">cc69239e629b14a03b5b87b9c5e8757c5c8f594b</guid><source url="https://aws.amazon.com/blogs/machine-learning/get-answers-from-tables-with-tabular-search-in-amazon-bedrock-managed-knowledge-base">AWS Machine Learning Blog</source></item><item><title>在 Amazon Bedrock 上为开源权重模型使用提示缓存</title><link>https://aws.amazon.com/blogs/machine-learning/use-prompt-caching-with-open-weight-models-on-amazon-bedrock</link><description>很多生成式 AI 应用每次请求都要重发相同的长上下文，提示缓存让 Bedrock 上的开源权重模型复用这部分内容，从而降低输入成本与延迟。</description><pubDate>Mon, 16 Nov 2026 15:42:41 GMT</pubDate><guid isPermaLink="false">7381db88317ca65963cc4aebf0b6d85f28f5015b</guid><source url="https://aws.amazon.com/blogs/machine-learning/use-prompt-caching-with-open-weight-models-on-amazon-bedrock">AWS Machine Learning Blog</source></item><item><title>LangChain：从任务型智能体走向目标型智能体</title><link>https://www.langchain.com/blog/from-tasks-to-goals</link><description>LangChain 正在把智能体从只做指定任务，转向能自己朝着开放目标推进的形态。</description><pubDate>Mon, 16 Nov 2026 23:53:29 GMT</pubDate><guid isPermaLink="false">cfa79461fd2cb8ae502f34b42790f5b5be9724cb</guid><source url="https://www.langchain.com/blog/from-tasks-to-goals">LangChain Blog</source></item><item><title>LangChain 提出目标工程：让智能体自己达成开放目标</title><link>https://www.langchain.com/blog/goal-engineering</link><description>未来的智能体要接手开放式目标并一路推进到底，LangChain 把这套做法叫做目标工程。</description><pubDate>Mon, 16 Nov 2026 23:53:29 GMT</pubDate><guid isPermaLink="false">b576b148cff59b595e9defbe6e267f2e2f5e4077</guid><source url="https://www.langchain.com/blog/goal-engineering">LangChain Blog</source></item><item><title>LangChain 谈 2026 年末上下文工程的新挑战</title><link>https://www.langchain.com/blog/context-engineering-challenges-late-2026</link><description>LangChain 的 CEO 分享了主动式智能体出现后，工程上在上下文管理方面遇到的最大难题。</description><pubDate>Mon, 16 Nov 2026 23:53:29 GMT</pubDate><guid isPermaLink="false">7a45c5f6802d9361ea14bcb772f3dfc597c32e17</guid><source url="https://www.langchain.com/blog/context-engineering-challenges-late-2026">LangChain Blog</source></item><item><title>LangChain 围绕智能体重建 LangSmith：一个记录管全部</title><link>https://www.langchain.com/blog/we-rebuilt-langsmith-around-the-agent</link><description>LangSmith 现在以单个智能体记录为中心，开发、运行、改进和治理都能在一个地方完成。</description><pubDate>Mon, 16 Nov 2026 18:37:11 GMT</pubDate><guid isPermaLink="false">fc9fb307ba1229e7c51196c752e62ad8596581cb</guid><source url="https://www.langchain.com/blog/we-rebuilt-langsmith-around-the-agent">LangChain Blog</source></item><item><title>图解 AI 智能体新书：概念与代码，教你用 LLM 搭建智能体</title><link>https://www.oreilly.com/library/view/an-illustrated-guide/9798341662681</link><description>O&#x27;Reilly 出了一本图文并茂的智能体指南，用概念讲解加代码示例带读者上手搭建 LLM 智能体。</description><pubDate>Tue, 17 Nov 2026 02:18:46 GMT</pubDate><guid isPermaLink="false">6081b5dff7bcafe1b366df0654695fa52d446be9</guid><source url="https://www.oreilly.com/library/view/an-illustrated-guide/9798341662681">Hacker News AI (hnrss)</source></item><item><title>智能体系统设计：自学路线图</title><link>https://www.kdnuggets.com/agentic-system-design-a-self-study-roadmap</link><description>这份七步自学路线图带你系统学习智能体系统设计，覆盖工具、记忆、控制循环、多智能体协作、评估与部署。</description><pubDate>Mon, 16 Nov 2026 13:00:00 GMT</pubDate><guid isPermaLink="false">454045de0d11fc348aeeddc9e257b3982fe41051</guid><source url="https://www.kdnuggets.com/agentic-system-design-a-self-study-roadmap">KDnuggets</source></item></channel></rss>
