每日简报

2026-11-09

← 历史归档

sherlock-project/sherlock

Python · ★ 104,056 · 🍴 12,160 · 📈 2,373 stars today

Hunt down social media accounts by username across social networks

中文介绍 Sherlock 是一个用户名搜索引擎,能够跨社交网络追踪某个用户名下注册的所有账号。它通过调用各平台公开 API 或页面,在数百个社交网站中批量检测账号存在性,常用于 OSINT 调查、安全审计及个人隐私暴露自查,纯 Python 实现,支持命令行和 Docker 部署。

dsx-ai-factory/nvke

Go · ★ 891 · 🍴 111 · 📈 345 stars today

NVIDIA Kubernetes Engine (NVKE). A management plane for GPU Kubernetes on bare metal.

中文介绍 NVKE 是 NVIDIA 推出的 Kubernetes 引擎,专为裸机 GPU 集群提供管理平面。它简化了在裸金属服务器上部署 GPU 工作负载的流程,解决传统 Kubernetes 对 GPU 资源调度复杂、驱动兼容性差等问题。适用于 AI 训练、推理及高性能计算场景,帮助运维人员快速搭建 GPU 云平台。

milind-soni/OpenMausBot

TypeScript · ★ 18,464 · 🍴 2,552 · 📈 1,439 stars today

Open-source Grok Bot alternative with a virtual machine that bots can use

中文介绍 OpenMausBot 是开源 Grok Bot 替代方案,其核心特色是内置虚拟机供机器人执行任务。它让开发者能编写和运行自定义机器人逻辑,在隔离环境中安全操作,适用于构建自动化对话机器人或智能体。项目面向希望自主掌控 AI 机器人行为的开发者,提供灵活的扩展机制。

anthropics/claude-quickstarts

Python · ★ 19,496 · 🍴 3,282 · 📈 32 stars today

A collection of projects designed to help developers quickly get started with building deployable applications using the Claude API

中文介绍 这是 Anthropic 官方的 Claude API 快速入门项目合集,包含多个可部署应用示例。每个项目演示如何用 Claude 构建实际功能,如聊天机器人、文档处理、代码生成等,帮助开发者快速上手。适合希望将 Claude 集成到自己产品中的工程师,附有完整代码和部署指南。

vercel-labs/agent-browser

Rust · ★ 47,733 · 🍴 3,266 · 📈 815 stars today

Browser automation CLI for AI agents

中文介绍 agent-browser 是 Vercel Labs 出品的浏览器自动化命令行工具,专为 AI Agent 设计。它让智能体能够通过简单命令控制真实浏览器,执行点击、填写、截图、抓取内容等操作,无需编写复杂脚本。适用于 AI 驱动的网页测试、数据抓取和自动任务执行,支持 Playwright 底层引擎。

meta/astryx

TypeScript · ★ 17,124 · 🍴 1,353 · 📈 875 stars today

An open source design system that's fully customizable and agent ready

中文介绍 Astryx 是 Meta 开源的设计系统,主打完全可定制并支持 AI Agent 集成。它提供一套可主题化的 UI 组件库,开发者可灵活调整视觉风格和交互行为,同时内置 Agent 友好的接口,便于 AI 助手操作界面。适合构建可定制化、未来面向智能交互的 Web 应用。

anthropics/skills

Python · ★ 189,760 · 🍴 22,355 · 📈 792 stars today

Public repository for Agent Skills

中文介绍 这是 Anthropic 官方公开的 Agent Skills 仓库,包含一系列预构建技能模块,供 Claude 及同类 Agent 使用。技能封装了特定领域知识或操作流程,如数据分析、文件处理、API 调用等,使 Agent 能即插即用地扩展能力。开发者可复用、贡献技能,加速智能体应用开发。

google-deepmind/mujoco

C++ · ★ 16,478 · 🍴 1,954 · 📈 41 stars today

Multi-Joint dynamics with Contact. A general purpose physics simulator.

中文介绍 MuJoCo 是 Google DeepMind 维护的通用物理仿真器,专注于多关节动力学与接触模型。它能够高效模拟机器人、生物力学系统及刚体交互,提供精确的接触力和约束求解。广泛应用于强化学习训练、机器人控制研究及动画物理效果,是科研和工业界常用的仿真平台。

ranxianglei/billion-context

TypeScript · ★ 5,251 · 🍴 409 · 📈 460 stars today

A context-compression plugin for month-long single sessions (billions of tokens) ,small context windows (a 100K context is enough) and token savings (5x fewer tokens) .上下文压缩插件,月级长会话会话, 兼顾小窗口(100k上下文足矣)省token(省5倍token) . billion-context is all you need

中文介绍 billion-context 是一个上下文压缩插件,专为超长会话场景设计。它通过压缩历史对话,使模型在仅需 100K 小窗口的情况下即可维持月级单会话,并减少约 5 倍 token 消耗。适用于长期运行 AI 助手或持续交互应用,有效降低推理成本并延长会话持续时间。

From Autonomous Agents to Cognitive Symbionts: Toward AI-Augmented Human Intelligence

👍 119

For a decade, progress in artificial intelligence has been measured by how much a machine can do on its own. Large foundation models now write code and run multistep tasks, and personal agents act on a person's behalf. Yet people still forget what they once knew, miss what they are not looking at, f

中文介绍 本文反思以「自主性」为尺度的AI进展,提出AI增强人类智能的新范式:AI作为「认知共生体」,与人类协作互补,而非单纯替代。作者主张重新定义进步标准,关注AI如何扩展而非取代人类认知能力,并探讨了实现该愿景的挑战与路径。

VFlash: Fast, Accurate Visual Planning via Flash Primitive Decoding

👍 107

Models for physical AI tasks, such as robotics and autonomous driving, need not reason solely in language. They can use visual plans--points, bounding boxes, and trajectories--to represent spatial goals and anticipate motion. However, autoregressive vision-language models generate these plans one co

中文介绍 VFlash提出一种基于Flash原语解码的快速视觉规划方法,用于机器人和自动驾驶等物理AI任务。传统自回归视觉语言模型逐点生成轨迹,效率低下。VFlash通过并行解码视觉原语(点、框、轨迹),大幅加速规划生成,同时保持高精度,显著提升了实时决策能力。

SLA: Scaling Spatial-Language-Action Model via Continual Pre-training

👍 93

Reliable robotic manipulation requires both semantic understanding and precise control in three-dimensional space. However, vision-language-action (VLA) foundations largely inherit visual representations from 2D vision-language pre-training, while spatial supervision is often introduced only during

中文介绍 SLA提出一种通过持续预训练扩展空间-语言-动作(VLA)模型的方法。现有VLA模型依赖2D视觉语言预训练特征,缺乏精确3D空间理解。SLA引入空间监督信号进行持续预训练,显著提升机器人在三维空间中的操作精度与语义理解,并在多项操作任务上刷新SOTA。

Decoupled Predictive Decision Making: A Shared Conditional Predictor for Agents

👍 81

Agent decisions depend on empirical relationships that deployment observations reveal and task descriptions alone may leave unresolved. We introduce Decoupled Predictive Decision Making (DPDM), an architecture that formulates heterogeneous agent decisions as conditional prediction with a shared pret

中文介绍 本文提出解耦预测决策(DPDM)架构,将异构智能体决策统一为条件预测问题,使用共享预训练预测器。该架构能利用部署时观测数据,解决任务描述中未明确的环境关系,提升决策泛化性与鲁棒性,为多智能体系统提供通用决策框架。

Harness-S: Learning from Sibling Harnesses for Generalizable Agent Training

👍 66

Harness evolution from a common base exhibits Locality, Complementarity, and Nonstationarity: variants repair different failures, and their utility changes as the policy learns. Rather than seeking a universally superior harness, we introduce Harness-S to selectively rescue failures of a base execut

中文介绍 Harness-S提出从兄弟训练框架(harness)中学习,以提升智能体训练的泛化能力。针对基线harness的失败模式,该方法有选择地修复特定缺陷,而非追求普遍优越的harness。实验表明,Harness-S能有效提升策略在未见任务上的表现,并降低训练成本。

Random Is Almost Enough? Stable and Low-Cost Benchmark Compression via BestRand

👍 51

Benchmark compression reduces the often prohibitive time and cost of model evaluation by evaluating models on a small benchmark subset that closely matches full-benchmark results. To quantify estimation error, we derive an accurate analytical approximation to the error distribution when estimating f

中文介绍 本文提出BestRand基准压缩方法,通过选择与全基准结果高度匹配的小子集,大幅降低模型评估的时间和成本。作者推导了估计误差的精确解析近似,并证明随机选择结合最优子集选择能达到稳定且低成本的压缩效果,优于现有方法。

WeLM-Image: A Unified Image Generation Model towards Visual Intelligence

👍 44

We present WeLM-Image, a unified image foundation model that integrates multimodal understanding into text-to-image generation, instruction-guided editing, and multi-reference synthesis. Designed for single-GPU deployment in WeChat's large-scale consumer applications, the model processes text, refer

中文介绍 WeLM-Image是一个统一图像基础模型,整合了多模态理解、文生图、指令引导编辑和多参考合成能力。该模型专为微信大规模消费级应用设计,支持单GPU部署,在多种图像生成与理解任务上展现了强大的性能与灵活性。

EvoMemState: Evolving Memory State for Personal Agents

👍 33

Long-term memory enables large language model (LLM) agents to provide continuous personalized assistance by retaining user information across interactions. Existing systems support this process through fact extraction, organization, updating, and retrieval, but largely maintain memories as independe

中文介绍 EvoMemState提出一种演化记忆状态机制,用于个人化LLM智能体的长期记忆管理。传统方法将记忆视为独立事实,而本方法通过状态演化,动态更新和关联用户信息,实现更连续、更个性化的服务。实验显示其能有效提升多轮交互中的记忆准确性与服务质量。

Halley: A Predictive Control Plane for Optical AI Data Center Networks

👍 33

A frontier AI model is trained in an AI data center, on tens of thousands of accelerators joined by a dedicated network, and every training step ends with all of them exchanging data at once. Between racks, that network is optical, and each of its four layers, from lasers to switches, runs a fixed o

中文介绍 本文提出Halley预测控制平面,用于光互连AI数据中心网络。针对AI训练中数千加速器同步通信的挑战,Halley预测网络流量需求,动态调整光网络层配置(激光器、交换机等),减少拥塞与等待,提升训练吞吐量,并降低能耗。

Open-HumanScale: An Open Recipe for Scaling Human and Robot Experience for Embodied Foundation Models

👍 32

Egocentric human video offers a scalable source of manipulation priors for robot learning, yet translating raw data scale into generalizable robot policies demands rigorous data curation and training recipes. We introduce Open-HumanScale, the first fully open recipe for scaling egocentric pretrainin

中文介绍 Open-HumanScale提出首个完全开源的具身基础模型预训练配方,用于扩展人类和机器人操作经验。通过严格的数据筛选和训练策略,将大规模第一视角人类视频转化为可泛化的机器人策略。该配方在模拟和真实机器人上均取得显著性能提升,并开源以促进复现。

VisSearch-Bench: Evaluating Long Horizon Visual Search in Multimodal Agents

👍 22

Multimodal agents use web and image tools to answer questions beyond their internal knowledge. Some search tasks use images mainly to identify an initial entity, then rely on text retrieval. These tasks offer limited insight into how agents connect information across images, objects, and search step

中文介绍 VisSearch-Bench是首个评估多模态智能体长程视觉搜索能力的基准。现有任务多依赖文本检索,而本基准要求智能体跨图像、物体和搜索步骤连接信息,测试其多步视觉推理与工具使用能力。实验揭示了当前多模态智能体在长程视觉搜索中的显著不足。

Towards Foundation Decision Models: Drex

👍 22

Autoregressive approaches with large language models, such as long chain-of-thought reasoning, have become the default for new AI applications because they reach higher accuracy on a wide range of tasks. In production, however, they give inconsistent answers, require parsing of outputs, and add cost

中文介绍 Drex提出一种基础决策模型,针对自回归LLM在生产环境中输出不一致、需要解析和成本高的问题。Drex采用非自回归架构,直接输出决策结果,避免长思维链的生成开销,同时保持高准确率。实验表明其在多种决策任务上性能接近或优于传统方法,且推理成本更低。

D3D: Decoupled Diffusion Drafting for Asynchronous Speculative Decoding

👍 20

Target-conditioned diffusion drafters predict candidate tokens in parallel, yet each new block must wait for updated target-model hidden states. This forces drafting and verification to run sequentially, limiting speedup even though each block is generated in one pass. We introduce D3D, a standalone

中文介绍 D3D提出解耦扩散草稿机制,用于异步推测解码。目标条件扩散草稿模型并行预测候选token,但需等待目标模型隐藏状态更新,导致串行瓶颈。D3D通过独立生成草稿块,实现草稿与验证异步执行,显著提升推测解码加速比,且不损失生成质量。

TAC-CoT: Tool-Empowered Implicit Chain-of-Thought via Interleaving Explicit Tokens

👍 17

LLM-based agentic systems are increasingly combined with CoT reasoning to solve tasks that require accurate reasoning, external knowledge, and multi-step deliberation. Yet, explicit CoT incurs substantial generation overhead, while implicit CoT, which replaces textual rationales with compact continu

中文介绍 TAC-CoT提出工具增强的隐式思维链方法,通过交错显式token来提升LLM智能体的推理能力。隐式CoT用紧凑连续向量替代文本推理,但缺乏工具调用能力。TAC-CoT在隐式CoT中插入工具调用token,兼顾推理效率与外部知识利用,在多步推理任务上实现更优性能与成本平衡。

Hybrid Computer Use: A Unified Programmatic, Semantic, and Visual Interface for General Agents

👍 17

Computer-use agents increasingly tackle tasks ranging from everyday desktop activities to long-horizon professional workflows. Yet interfaces centered on screenshots and mouse-and-keyboard actions can impose a human-interaction prior, requiring repeated screen interpretation while leaving the comput

中文介绍 本文提出混合计算机使用界面,统一程序化、语义和视觉控制,用于通用智能体。传统截图和鼠标键盘交互依赖人类交互先验,效率低。新界面允许智能体直接调用程序功能、操作语义元素和视觉组件,提升长程任务执行的成功率与速度,适用于桌面和专业工作流。

On the Opportunities and Challenges of Prefill-to-Autoregressive Cross-Model Transfer

👍 15

Prefill-to-Autoregressive Cross-model Transfer (PACT), a form of latent communication between language models, combines a larger model's prompt representations with a smaller model's efficient generation, offering a promising way to balance capability and inference cost. However, existing systems pr

中文介绍 本文探讨预填充到自回归跨模型迁移(PACT)的机遇与挑战。PACT将大模型的提示表示与小模型的高效生成结合,平衡能力与推理成本。作者系统分析现有系统的不足,并提出改进方向,为跨模型潜在通信技术提供理论框架和实验证据。

Why Pay for Long Reasoning When Rewrite Do Trick? Causal Effects of Reasoning History Compaction on Future Generation and Agent Cost

👍 15

Reasoning substantially improves language models' ability to solve difficult problems, but long reasoning traces make these gains expensive. Efficient reasoning should express useful analysis in fewer tokens by reducing repetition and text that adds little value to solving the problem. Progress in p

中文介绍 本文研究推理历史压缩对未来生成和智能体成本的因果效应。长推理链虽提升准确率但成本高昂。作者提出通过重写压缩冗余推理内容,发现压缩历史能降低后续生成成本,且不影响性能。因果分析表明,去除低价值推理token可显著提升效率。

Terminal-style Mathematical Modeling Bench: The First Agentic and Verifiable Benchmark for Mathematical Modeling

👍 11

Mathematical modeling turns an open-ended real-world problem into assumptions, a model, a solution, and a defended recommendation, and its answers matter only if they are both close to a credible reference and physically feasible; existing benchmarks either score solver fragments or grade reports wi

中文介绍 本文提出Terminal-style数学建模基准,首个智能体化且可验证的数学建模评估。与现有评分片段或报告不同,该基准要求智能体产出完整建模流程,包括假设、模型、求解和论证,并验证结果与参考解的一致性及物理可行性。测试显示当前智能体在此类任务上表现不佳。

The SIMPLE Law for Energy Cost of LLM Inference

👍 11

Generative AI is billed per token, yet tokens are not equally expensive. Each token enters the KV-cache and raises the cost of those that follow, so energy grows non-linearly along a sequence whose tokens look identical. No analytical model states what a token costs, and the most accurate SOTA model

中文介绍 本文提出SIMPLE定律,解析描述LLM推理的能耗成本。现有按token计费忽略了token间的依赖(KV缓存导致非线性能耗增长)。SIMPLE定律提供精确分析模型,量化每个token的实际成本,为优化推理策略和定价机制提供理论依据,实验验证其准确性。

Decode Before You Hear: Speculative Listening for Chunk-Interleaved Full-Duplex Speech Language Models

👍 10

Chunk-interleaved full-duplex speech language models listen and speak in fixed time chunks, and each chunk must be processed before the next one arrives. Their workload is asymmetric across chunks: listen chunks leave compute idle while awaiting audio, and speak and tool-call chunks require serial g

中文介绍 本文提出推测性监听(Speculative Listening)技术,用于块交错全双工语音语言模型。针对监听块计算空闲、说话和工具调用块串行瓶颈的问题,该方法在监听块内提前解码未来说话内容,实现计算复用。实验表明能显著降低响应延迟,提升全双工交互效率。

How I Turn One Image Into a 3D Simulator That Takes Itself Apart

@Nyxera0x · 22.1K 粉丝 · 199.5K 阅 · 509 赞 · 66 转

I gave Claude a job description instead of a prompt. Now one picture becomes a 3D build in code that Claude checks against the picture and films taking itself apart. Below: the brief, the checklist

中文介绍 用一份职位描述代替 prompt,让 Claude 将一张图片生成为可自动拆解的 3D 构建代码,并通过自检机制确保其与图片一致。作者展示了如何用任务简报驱动 AI 完成复杂视觉创作。

What does it actually take to make an AI Financial Assistant good?

@swangentr · 54.7K 粉丝 · 105.9K 阅 · 513 赞 · 42 转

@dubinvest just launched Arlo, our first AI financial assistant. We've come a long way from the first version, which was a single prompt that grew toward 200k tokens... Our engineering team

中文介绍 AI 金融助手 Arlo 的开发复盘:从最初的单个 prompt 膨胀到近 20 万 token,工程团队分享了优化过程,以及如何构建一个真正靠谱的金融助手所涉及的架构与体验设计。

Growing Together with Our Plugin Community

@DeepSeekHarness · 81.6K 粉丝 · 85.3K 阅 · 1.5K 赞 · 127 转

Since DeepSeek Harness launched, more and more people have been working and exploring in it, and more and more developers have been extending what it can do through plugins. Today we'd like to share

中文介绍 DeepSeek Harness 插件生态社区更新:官方回顾了上线以来的进展,越来越多的开发者通过插件扩展工具能力,并分享了优秀插件案例及贡献方式。

The ultimate guide to Grok Bot for creators

@XCreators · 10.8M 粉丝 · 58.4K 阅 · 690 赞 · 55 转

By @Ethan_Ng_13 Being a creator is way more than filming and posting a video. A single 60-second short can take days. There's the idea, the research, the script, the shoot, the edit, the post, the

中文介绍 面向创作者的全流程 Grok Bot 指南:从创意、调研、脚本到拍摄剪辑,一条 60 秒短视频往往耗费数天。作者系统梳理了 Grok Bot 如何贯穿并提效整个创作链路。

Securing AI Agents: Lessons From Real-World Deployment

@UberEng · 71.7K 粉丝 · 41.5K 阅 · 519 赞 · 66 转

Introduction Agentic AI introduces a fundamentally different security model from traditional software. Unlike conventional services that execute predefined logic, agents dynamically reason over

中文介绍 Uber 工程团队分享 AI Agent 安全实战:与传统软件不同,Agent 需动态推理,安全模型需重构。文章提炼了真实部署中的风险维度与防护经验,为 Agent 安全提供参考。

Claude code/Codex+Seedance AI Video: The Full Process from an Idea to a Finished Film

@EianLu · 30.0K 粉丝 · 38.7K 阅 · 529 赞 · 50 转

This article solves only one problem: how to rely on a multi-model workflow and the things to watch out for during actual operation to make an AI video you are satisfied with, using as little money as

中文介绍 分享用 Claude Code、Codex 加 Seedance 的多模型工作流,以低成本完成从想法到成片的 AI 视频全流程,并总结了实际运行中的注意事项。

Your coding agents made you a manager

@addyosmani · 491.5K 粉丝 · 36.8K 阅 · 721 赞 · 70 转

The skills for managing agents are pretty much the skills for managing a team, and you don't need to be a manager to learn them. If you're spending all day with a coding agent, you are managing.

中文介绍 核心观点:管理 AI 编码 Agent 所需技能与管理团队几乎相同。作者认为日常使用 Agent 的开发者实际上已在扮演管理者角色,应主动学习相关管理方法。

The Genesis of $IMPRINT

@imprint_bnb · 2.6K 粉丝 · 9.2K 阅 · 728 赞 · 548 转

Agents will paint it. A 100x100 canvas on BNB Chain, open for 12 hours, where only .ai agents can join. When the clock runs out, the canvas freezes and becomes the logo of $IMPRINT. Half the supply

中文介绍 $IMPRINT 代币创世事件:BNB Chain 上开出一张 100x100 画布,限时 12 小时,仅限 .ai 代理参与绘画。时间截止后画作冻结成为代币 LOGO,一半供应量将由此分配。

How Stanford Teaches AI-Native Engineering — Mihail Eric, Monaco

What happens when students can one-shot last year’s homework? Inside CS146S and the engineering team at Monaco.

中文介绍 斯坦福大学开设CS146S课程,教授AI原生工程。课程由Mihail Eric与Monaco团队合作,探讨当学生能通过一次提示完成去年作业时,如何设计课程内容。

[One Million AI Engineer Special] How Stanford Teaches AI-Native Engineering — Mihail Eric, Monaco

What do you assign when a coding agent can finish last year’s homework in one prompt? In the second run of Stanford’s CS146S, Mihail Eric brought open-source maintainers into the course. Students have to learn an unfamiliar repo, propose a change, and ...

中文介绍 斯坦福CS146S第二期课程引入开源维护者,学生需学习不熟悉的代码库、提出修改建议。课程旨在培养学生应对AI代理能快速完成任务的新环境。

[AINews] Anthropic IPO: Whither Now?

November 6, 2026 — ARC Prize tests Inkling 2 Small, and Thinking Machines explains how it caught early checkpoints bypassing guardrails.

中文介绍 2026年11月6日AI新闻:Anthropic IPO前景受关注;ARC Prize测试Inkling 2 Small模型;Thinking Machines披露其发现早期检查点绕过安全护栏的问题。

Controlling Biology - Fabian Theis, Helmholtz Munich

As systems biology efforts heat up to map biology at higher resolution, greater breadth and more modalities, Fabian Theis is pushing the frontier from mapping to predicting to controlling biology

中文介绍 系统生物学研究升温,Helmholtz Munich的Fabian Theis正推动从生物图谱绘制向预测和控制生物学方向拓展,提升分辨率、广度和模态覆盖。

I watched a humanoid robot clean my apartment. It felt like a taste of the future.

I’m not the sort of person who typically pays for a house cleaner. I’m too frugal, and I kind of enjoy putting on a podcast and going to town on my tub with a Magic Eraser. But when I saw an X post about a new humanoid-robot cleaning service in San Francisco, I immediately signed…

中文介绍 MIT科技评论记者体验旧金山的人形机器人清洁服务,认为其代表未来趋势。作者描述机器人在公寓内完成清洁任务的过程,感觉像未来的预演。

Partnering with Hacktron to find vulnerabilities in OpenAI’s systems

OpenAI is partnering with Hacktron to find vulnerabilities, test connections with third-party services, and strengthen defenses with frontier AI.

中文介绍 OpenAI与Hacktron合作,寻找系统漏洞,测试与第三方服务的连接,并利用前沿AI加强防御能力。

Measuring how AI agents handle everyday wet-lab science

OpenAI introduces BioBasicsBench, a 100-task benchmark measuring how well AI agents interpret raw experimental data and handle everyday wet-lab science.

中文介绍 OpenAI发布BioBasicsBench基准,包含100项任务,用于评估AI代理如何解析原始实验数据并处理日常湿实验科学工作。

Cult.fit builds operational muscle with Codex

Cult.fit is using Codex to help teams across functions build tools, move faster, and change how work gets done.

中文介绍 印度健身公司Cult.fit使用OpenAI Codex帮助跨部门团队构建工具、加快工作流程,并改变工作方式。

Australian universities are creating AI-native graduates

Australian universities are expanding access to ChatGPT and bringing AI into teaching and learning to prepare students for an AI-enabled workforce.

中文介绍 澳大利亚多所大学扩大ChatGPT使用范围,将AI融入教学,培养适应AI驱动职场的毕业生。

GPT‑6 Cyber: Advancing the frontier of aligned cybersecurity agents

Introducing GPT‑6 Cyber, OpenAI’s cybersecurity model for finding exploitable vulnerabilities and developing fixes, trained to stay within authorized scope.

中文介绍 OpenAI推出GPT-6 Cyber网络安全模型,专门用于发现可利用漏洞并开发修复方案,且训练中严格限定在授权范围内操作。

市场总览

美股整体技术面偏弱:SPY近5日跌2.7%,RSI 35.1接近超卖;QQQ近5日跌4.98%,RSI 31.4更低;NVDA RSI 30.8且MACD负柱扩大。中概股相对强势:JD近5日涨13.36%且RSI 72超买,BABA与0700.HK均现MACD金叉。加密市场承压,恐慌贪婪指数29显示恐慌,总市值24小时跌4.68%;BTC RSI 36.5、ETH RSI 34均偏弱,SOL RSI 28.3超卖。商品外汇分化:黄金期货MACD金叉但价格仍低于SMA50;原油期货近5日跌5.71%且MACD死叉;美元指数多头排列。宏观方面,VIX近5日大涨32.03%至25.64,显示市场波动加剧;10Y美债收益率RSI 36.9低于50。整体来看,风险资产多数处于下行或中性震荡,中概股出现独立反弹迹象。

今日关注

JD 京东 (JD)
偏上行

京东近5日上涨13.36%,RSI 14达72处于超买区,MACD柱0.244为正且MACD线0.244高于信号线-0.1586形成金叉,股价29.11突破SMA20 26.56和SMA50 26.88,短期动能强劲,但超买需留意回调风险。

GOOGL Alphabet (GOOGL)
偏下行

Alphabet股价284.78低于SMA20 317.44、SMA50 332.93及SMA200 339.51呈空头排列,RSI 14为29.9处于超卖区,MACD线-16.58低于信号线-12.79且柱为负,且SMA50已下穿SMA200形成死叉,下行趋势明确。

SOL-USD Solana (SOL-USD)
偏下行

Solana现价87.66低于SMA20 97.28和SMA50 107.23,RSI 14为28.3进入超卖区,MACD线-4.78低于信号线-4.14,柱状图为负,近5日下跌5.66%,且距52周高点回撤48.92%,下行压力显著。

BABA 阿里巴巴 (BABA)
中性

阿里巴巴现价109.45略高于SMA50 108.29但低于SMA200 124.64,RSI 14为57.7中性偏强,MACD线-1.26高于信号线-2.23且柱状图转正,近5日上涨11.64%显示反弹,但长期均线仍压制,多空信号交织。

DX-Y.NYB 美元指数 (DXY)
偏上行

美元指数现价102.17高于SMA20 101.95、SMA50 101.14及SMA200 99.73呈多头排列,RSI 14为56.6中性偏强,MACD线0.16高于信号线0.21但柱状图为正,且接近52周高点,中期上升趋势完好。

全部资产

^VIX

VIX 恐慌指数

$25.64 -9.01%
5 日
+32.03%
距 52w 高
-27.4%
RSI(14)
63.2
趋势
中性
SMA 20 / 50 / 200
18.83 / 16.92 / 18.40
MACD / 信号
1.853 / 1.074

^TNX

10Y 美债收益率 (%)

$4.86 +0.45%
5 日
-1.42%
距 52w 高
-9.4%
RSI(14)
36.9
趋势
中性
SMA 20 / 50 / 200
5.03 / 5.03 / 4.56
MACD / 信号
-0.058 / -0.031
接近 52 周低

DX-Y.NYB

美元指数 DXY

$102.17 -0.13%
5 日
+0.41%
距 52w 高
-0.8%
RSI(14)
56.6
趋势
多头
SMA 20 / 50 / 200
101.95 / 101.14 / 99.73
MACD / 信号
0.163 / 0.213
接近 52 周高多头排列

SPY

S&P 500 ETF

$735.29 +0.31%
5 日
-2.70%
距 52w 高
-5.9%
RSI(14)
35.1
趋势
中性
SMA 20 / 50 / 200
763.36 / 765.68 / 730.38
MACD / 信号
-6.076 / -2.652

QQQ

Nasdaq 100 ETF

$656.10 +1.40%
5 日
-4.98%
距 52w 高
-14.0%
RSI(14)
31.4
趋势
中性
SMA 20 / 50 / 200
707.64 / 720.97 / 680.89
MACD / 信号
-16.635 / -9.993

AAPL

Apple

$342.37 -0.52%
5 日
-0.66%
距 52w 高
-3.7%
RSI(14)
51.5
趋势
多头
SMA 20 / 50 / 200
343.38 / 336.01 / 298.81
MACD / 信号
2.067 / 2.835
多头排列

MSFT

Microsoft

$530.61 -0.99%
5 日
+1.67%
距 52w 高
-5.1%
RSI(14)
52.1
趋势
多头
SMA 20 / 50 / 200
531.41 / 516.65 / 439.85
MACD / 信号
4.371 / 4.957
多头排列

NVDA

Nvidia

$186.64 +2.78%
5 日
-8.80%
距 52w 高
-23.3%
RSI(14)
30.8
趋势
中性
SMA 20 / 50 / 200
211.55 / 219.81 / 204.67
MACD / 信号
-8.238 / -5.041

GOOGL

Alphabet

$284.78 +0.78%
5 日
+1.77%
距 52w 高
-30.3%
RSI(14)
29.9
趋势
空头
SMA 20 / 50 / 200
317.44 / 332.93 / 339.51
MACD / 信号
-16.577 / -12.786
死叉(SMA50↓SMA200) (6 天前)RSI 超卖空头排列

TSLA

Tesla

$358.22 +4.23%
5 日
-10.40%
距 52w 高
-28.2%
RSI(14)
42.4
趋势
空头
SMA 20 / 50 / 200
380.42 / 372.14 / 383.87
MACD / 信号
1.464 / 5.689
MACD 死叉 (1 天前)空头排列

META

Meta

$610.15 +0.84%
5 日
-0.79%
距 52w 高
-21.8%
RSI(14)
36.7
趋势
中性
SMA 20 / 50 / 200
662.67 / 674.92 / 631.94
MACD / 信号
-23.690 / -17.813
加密恐慌贪婪
29
恐慌
加密总市值
$2.50 T
-4.68% / 24h
BTC 主导率
60.3%
ETH 10.4%
24h 成交量
$52.6 B
活跃币 22,869

BTC-USD

Bitcoin

$75,074.35 -0.89%
5 日
-2.94%
距 52w 高
-30.1%
RSI(14)
36.5
趋势
中性
SMA 20 / 50 / 200
78,991.42 / 81,748.62 / 73,172.14
MACD / 信号
-1,593.374 / -1,246.640

ETH-USD

Ethereum

$2,141.79 -4.23%
5 日
-5.76%
距 52w 高
-41.4%
RSI(14)
34.0
趋势
中性
SMA 20 / 50 / 200
2,310.21 / 2,481.37 / 2,154.61
MACD / 信号
-66.899 / -61.914
MACD 死叉 (今天)

SOL-USD

Solana

$87.66 -2.24%
5 日
-5.66%
距 52w 高
-48.9%
RSI(14)
28.3
趋势
中性
SMA 20 / 50 / 200
97.28 / 107.23 / 89.00
MACD / 信号
-4.782 / -4.144
RSI 超卖

BABA

阿里巴巴 (BABA)

$109.45 +1.28%
5 日
+11.64%
距 52w 高
-39.6%
RSI(14)
57.7
趋势
中性
SMA 20 / 50 / 200
104.99 / 108.29 / 124.64
MACD / 信号
-1.264 / -2.233
MACD 金叉 (2 天前)

PDD

拼多多 (PDD)

$83.03 -1.10%
5 日
+5.62%
距 52w 高
-40.3%
RSI(14)
58.2
趋势
中性
SMA 20 / 50 / 200
80.53 / 79.82 / 90.46
MACD / 信号
0.554 / 0.049
MACD 金叉 (4 天前)

JD

京东 (JD)

$29.11 +1.64%
5 日
+13.36%
距 52w 高
-15.1%
RSI(14)
72.0
趋势
中性
SMA 20 / 50 / 200
26.56 / 26.88 / 28.55
MACD / 信号
0.244 / -0.159
MACD 金叉 (4 天前)RSI 超买

0700.HK

腾讯控股 (0700.HK)

HK$421.80 +0.48%
5 日
+7.55%
距 52w 高
-36.3%
RSI(14)
56.3
趋势
中性
SMA 20 / 50 / 200
403.30 / 421.38 / 473.26
MACD / 信号
-2.863 / -7.982
MACD 金叉 (4 天前)

GC=F

黄金期货

$4,129.60 +1.34%
5 日
+4.03%
距 52w 高
-26.1%
RSI(14)
51.4
趋势
空头
SMA 20 / 50 / 200
4,077.66 / 4,222.42 / 4,497.24
MACD / 信号
-43.367 / -60.911
MACD 金叉 (2 天前)空头排列

CL=F

WTI 原油期货

$90.95 +0.13%
5 日
-5.71%
距 52w 高
-23.9%
RSI(14)
42.4
趋势
中性
SMA 20 / 50 / 200
97.87 / 95.62 / 87.31
MACD / 信号
-0.743 / 0.744

USDCNY=X

美元 / 人民币

¥6.67 -0.21%
5 日
+0.20%
距 52w 高
-7.2%
RSI(14)
44.8
趋势
空头
SMA 20 / 50 / 200
6.68 / 6.70 / 6.79
MACD / 信号
-0.008 / -0.010
MACD 金叉 (1 天前)接近 52 周低空头排列
风险提示

本报告仅供技术指标解读参考,不构成投资建议。过去走势不代表未来表现,技术指标可能存在滞后或失效风险。市场受突发事件影响可能瞬间改变技术形态,请谨慎对待。

Moldova launches new power line to Romania to cut reliance on Russia

Moldovan President Maia Sandu says 'two more power lines to Romania are on the way'.

中文摘要 摩尔多瓦总统桑杜宣布,连接该国与罗马尼亚的新输电线路正式启用,以降低对俄罗斯电力的依赖,并称另外两条通往罗马尼亚的输电线路正在建设中。

Drone attack on Iraqi Kurdistan gas field triggers mass blackout

A drone attack has halted production at the Khor Mor gas field in northern Iraq, causing widespread power outages.

中文摘要 伊拉克北部库尔德地区霍尔莫尔气田遭无人机袭击,生产中断,导致当地大范围停电。

Pope Leo delivers stinging rebuke of Milei’s austerity measures during Argentina visit

Pontiff tells Argentina’s president the economy must benefit all on second stop of 12-day Latin America tour Pope Leo XIV has delivered a stinging critique of Argentine president Javier Milei’s harsh austerity measures, insisting growth must benefit all in a speech met with rapturous applause in Bue

中文摘要 教皇利奥十四世在访问阿根廷期间发表讲话,严厉批评总统米莱的紧缩措施,强调经济增长必须惠及所有人。这是其为期12天拉丁美洲之行的第二站。

‘I never executed Jesse Baird’: Beau Lamarre-Condon reenacts shooting when giving evidence in NSW trial

Former police officer has pleaded not guilty to murdering Baird and Luke Davies in Paddington in early 2024 Warning: This article contains graphic content Get our breaking news email, free app or daily news podcast Beau Lamarre-Condon has told a Sydney court he did not “execute” Jesse Baird as he ac

中文摘要 前警察博·拉马尔-康登在悉尼法庭受审时否认“处决”了杰西·贝尔德,并当场重现枪击过程。他对2024年初谋杀贝尔德及卢克·戴维斯的指控不认罪。

Polluted Delhi Tries ‘Green’ Fireworks to Clear the Air at Diwali

India’s top court decreed that only more environmentally friendly fireworks could be set off this weekend, citing air quality concerns. But environmental activists call it a “green wash.”

中文摘要 印度最高法院裁定,排灯节期间只允许燃放更环保的烟花,以改善德里严重的空气质量。但环保人士批评此举是“漂绿”。

Wagga Wagga family members of alleged DV murderer charged with being accessories after the fact

Danielle Madgwick died in regional NSW in September after suffering multiple stab wounds to her upper body Get our breaking news email, free app or daily news podcast Family members of a man accused of the domestic violence murder of a 23-year-old woman in south-western New South Wales have been cha

中文摘要 澳大利亚新南威尔士州沃加沃加一名23岁女子丹妮尔·马德威克遭家庭暴力谋杀。被控男子的多名家人因事后协助而面临指控。

Man charged with murder over fatal stabbing of HungryPanda delivery rider in Melbourne

Footscray man charged over death of 47-year-old Feng Jianming in the CBD on Friday night Get our breaking news email, free app or daily news podcast A man has been charged with murder after a delivery rider was stabbed to death while working, leaving his family bereft and a city in mourning. On Frid

中文摘要 澳大利亚墨尔本,一名47岁华人送餐员冯建明在市中心送餐时被刺死。一名男子被控谋杀,警方已展开调查。

Nvidia opens investigation after report CEO Jensen Huang was target of foreign intelligence

CEO reportedly ‌had close relationship ‌with a Chinese businesswoman who passed information to ‌Beijing Nvidia said ⁠on Sunday it “is initiating an independent investigation in light of recent reporting that founder and CEO Jensen Huang has been the target of a foreign intelligence effort”. The New

中文摘要 英伟达宣布,就创始人兼CEO黄仁勋涉嫌成为外国情报机构目标的报道启动独立调查。此前报道称,他与一名向北京传递信息的中国女商人关系密切。

India’s night lights up for Diwali

India is celebrating Diwali with prayers, lights and fireworks marking one of the country's biggest Hindu festivals.

中文摘要 印度庆祝排灯节,民众通过祈祷、点灯和放烟花庆祝这一重要的印度教节日,全国夜空被照亮。

Jeffries backs Hegseth impeachment as Democrats vow probes of Trump allies

Democratic lawmakers are mapping out their post-midterm agenda with pledges to lead committee investigations into Trump.

中文摘要 美国民主党众议员杰弗里斯支持对赫格塞思的弹劾,民主党人誓言在中期选举后对特朗普盟友展开委员会调查。

Iran war live: Trump to talk to Xi, Putin about Tehran, ‘denuclearization’

Trump says Iran land blockade is now in effect as Tehran threatens to 'rain down hell' if US resumes attacks.

中文摘要 特朗普宣布对伊朗实施陆地封锁,并称将与中国国家主席习近平和俄罗斯总统普京讨论伊朗“无核化”问题。伊朗威胁若美国恢复袭击将“降下地狱之火”。

Nvidia CEO’s affair sparks Chinese espionage probe

Nvidia CEO Jensen Huang has admitted to an affair with a woman suspected of spying for China.

中文摘要 英伟达CEO黄仁勋承认与一名涉嫌为中国从事间谍活动的女性有婚外情,该事件引发对中国间谍活动的调查。

Mexican mayor found alive after hours-long abduction

Rio Bravo mayor freed by captors after being abducted in Zacatecas state, known for cartel-linked violence.

中文摘要 墨西哥里奥布拉沃市长在萨卡特卡斯州遭绑架数小时后获释。该地区以贩毒集团相关暴力闻名。

Chancellor Merz to nominate Aigner as Germany’s first woman president

Ilse Aigner, speaker of the Bavarian state parliament, belongs to the Christian Social Union in Bavaria (CSU) party.

中文摘要 德国总理默茨将提名巴伐利亚州议会议长伊尔塞·艾格纳为总统候选人,她来自基社盟(CSU)。若当选,她将成为德国首位女性总统。

Foreign Funds Pile Into India as AI Trade Falters

Including Friday’s provisional figures, foreign funds were net buyers of close to $5 billion of stocks last week.

中文摘要 外资上周净买入印度股票近50亿美元,因人工智能交易表现疲软,资金转向印度市场。包括周五的初步数据,外资连续成为净买家。

China’s Cheap Domestic Gas Curbs LNG Buying as Winter Nears

China’s domestic natural gas prices are weakening, curbing the world’s biggest importer’s appetite for expensive seaborne supplies and potentially freeing up cargoes for Europe and other Asian markets during the peak winter demand season.

中文摘要 中国国内天然气价格走弱,抑制了全球最大进口国对昂贵海运LNG的购买意愿,可能在冬季需求高峰期释放货源给欧洲和其他亚洲市场。

Philippine Economy Unexpectedly Slows, Pressuring Marcos

The Philippines’ economic growth, one of Asia’s worst-performing, unexpectedly slowed for a fifth straight quarter on weak farm output, with high energy costs and El Niño continuing to cloud the outlook.

中文摘要 菲律宾经济增速连续第五个季度意外放缓,为亚洲表现最差之一。农产品产出疲弱、能源成本高企及厄尔尼诺现象继续拖累前景,给马科斯政府带来压力。

Mizuho Pierces Nomura’s Japan Bubble With Big-Ticket Global IPOs

Mizuho Financial Group Inc. is carving out a lucrative niche in Japan’s equity capital market by bridging the ravenous US demand for artificial intelligence fundraising and its home country’s deep pool of retail savings.

中文摘要 瑞穗金融集团通过连接美国对AI融资的旺盛需求与日本庞大的零售储蓄,在日本股票资本市场开辟了利润丰厚的细分市场,超越野村证券。

India Short Bonds May Underperform on Rate Hikes, Cash Squeeze

India’s shorter bonds will likely resume their underperformance against longer peers as tightening liquidity and monetary conditions rein in appetite for the rate-sensitive notes.

中文摘要 印度短期债券可能恢复对长期债券的跑输表现,因流动性收紧和货币条件趋紧抑制了对利率敏感债券的需求。

Lula Vows Fiscal Responsibility in First Address Since Election

President Luiz Inácio Lula da Silva pledged to preserve Brazil’s public finances without abandoning social programs in his first televised address since winning a fourth term, as he sought to reassure investors.

中文摘要 巴西总统卢拉在赢得第四个任期后的首次电视讲话中承诺,在不放弃社会项目的前提下维护公共财政,旨在安抚投资者。

Berkshire Plans First Japan Bond Offering Since April

Berkshire Hathaway Inc. has hired banks for its first yen bond sale since April as it prepares to raise more funds in Japan.

中文摘要 伯克希尔·哈撒韦已聘请银行筹备自4月以来首次日元债券发行,旨在日本筹集更多资金。

Nvidia launches investigation into CEO's affair with suspected Chinese spy

Jensen Huang also admitted to having been "unfaithful" to his wife and says he will cooperate with a security investigation.

中文摘要 英伟达就CEO黄仁勋与疑似中国间谍的婚外情展开调查。黄承认对妻子不忠,并表示将配合安全调查。

Doing these things now will help you save money at Christmas

Small decisions made now can make a big difference to your festive finances, charities say, so here are some tips.

中文摘要 慈善机构建议,现在做出一些小决定可以为节日开支节省大量资金,为圣诞节做好准备。

Sudden Distress at Top-Rated Firm Tests China’s Rating Crackdown

Less than six months ago, China’s central bank made its biggest push yet to tackle inflated credit ratings in the nation’s $5.6 trillion corporate debt market. Now it has a live test case.

中文摘要 中国央行不到六个月前对5.6万亿美元公司债市场的虚高信用评级发起最大规模整治,如今一家高评级公司突发困境,成为整治的试金石。

Gold Gains as Oil Extends Fall and Nvidia Probe Rattles Markets

Gold rose, erasing Friday’s decline, as oil extended a fall and an investigation into Nvidia Corp.’s chief executive officer weighed on technology stocks.

中文摘要 黄金上涨,抹去周五跌幅,因油价延续跌势,且针对英伟达CEO的调查打压科技股。

该源今日无内容。