TypeScript · ★ 4,712 · 🍴 196 · 📈 1,430 stars today
Next generation e2e testing framework for web and mobile apps.
中文介绍 面向 Web 与移动应用的下一代端到端测试框架，目标是统一两类平台的 e2e 测试流程。推测其提供跨端一致的测试 API 与运行器，减少浏览器与原生 App 各维护一套测试栈的成本。适合需要同时覆盖 Web 和移动端、并在 CI 中跑回归的测试团队。
TypeScript · ★ 96,612 · 🍴 8,520 · 📈 534 stars today
Persistent Context Across Sessions for Every Agent – Captures everything your agent does during sessions, compresses it with AI, and injects relevant context back into future sessions. Works with Claude Code, OpenClaw, Codex, Gemini, Hermes, Copilot, OpenCode + More
中文介绍 为 AI Agent 提供跨会话的持久记忆：自动记录 Agent 在一次会话中的全部操作，用 AI 压缩成摘要，并在后续会话中注入相关上下文。支持 Claude 等编码助手，解决长任务中断后上下文丢失、反复解释背景的问题，适合长期迭代同一项目的开发者。
Python · ★ 17,382 · 🍴 1,768 · 📈 456 stars today
Give your agent CAD superpowers.
中文介绍 给 AI Agent 加上 CAD 能力：让模型通过文本描述生成或修改三维 CAD 模型。推测其以工具或插件形式接入 Agent，把自然语言转成建模指令。面向需要快速产出概念模型的设计师、机械工程师，以及做 CAD 自动化的开发者。
TypeScript · ★ 25,591 · 🍴 6,593 · 📈 487 stars today
中文介绍 pingdotgg 团队（T3 生态）发布的代码相关项目，仓库暂未提供英文描述，具体功能与实现方式需查看 README 与源码。从命名看大概围绕 AI 辅助编码或代码生成工具，适合关注 T3 技术栈和 AI 编程工具的开发者跟进。
C++ · ★ 4,918 · 🍴 367 · 📈 994 stars today
Tool for automatic PS5 executables porting to Linux and Windows
中文介绍 把 PS5 可执行文件自动移植到 Linux 和 Windows 的工具。通过解析 PS5 平台二进制并转换为桌面系统可运行的形式，降低逐个人工适配的成本。主要面向主机模拟、逆向工程与游戏移植方向的研究者，用于实验性运行 PS5 程序。
Python · ★ 91,849 · 🍴 8,062 · 📈 1,156 stars today
Give your AI agent eyes to see the entire internet. Read & search Twitter, Reddit, YouTube, GitHub, Bilibili, XiaoHongShu — one CLI, zero API fees.
中文介绍 给 AI Agent 装上“眼睛”：一个 CLI 即可读取与搜索 Twitter、Reddit、YouTube、GitHub、Bilibili、小红书等平台内容，且不依赖付费 API。解决 Agent 难以联网获取社交与视频平台信息的问题，适合做舆情监控、内容聚合和调研类 Agent 的开发者。
Python · ★ 63,993 · 🍴 8,124 · 📈 758 stars today
World's first open-source, agentic video production system. 12 production pipelines, 100+ tools, 700+ agent skill and production-knowledge files. Turn your AI coding assistant into a full video production studio.
中文介绍 开源、Agent 驱动的视频制作系统，内置 12 条制作流水线、100 多个工具以及 700 多个 Agent 技能与制作知识文件。它把 AI 编码助手变成完整的视频生产链路，覆盖脚本到成片流程。适合想用 Agent 自动化产出短视频或营销视频的个人与团队。
Go · ★ 77,087 · 🍴 5,073 · 📈 526 stars today
Fast and extensible multi-platform HTTP/1-2-3 web server with automatic HTTPS
中文介绍 用 Go 编写的多平台 Web 服务器，支持 HTTP/1、HTTP/2、HTTP/3，以配置简单和默认自动申请、续期 HTTPS 证书著称。相比传统服务器省去手写证书与繁琐配置，常被用作反向代理和静态站点托管，适合自建服务的开发者与运维人员。
JavaScript · ★ 4,157 · 🍴 651 · 📈 1,444 stars today
Self-hosted gym & body-weight tracker — plan routines, log workouts (supersets, warm-ups, cardio), see which muscles are trained, fatigued or detrained, import from FitNotes/Strong/Hevy, passkey login. Your data, your server.
中文介绍 自托管的健身与自重训练记录工具：可制定训练计划，记录超级组、热身与有氧等训练内容，并查看各肌群处于训练、疲劳还是退化状态。支持从 FitNotes、Strong、Hevy 导入数据，提供 passkey 登录。适合想自己掌控训练数据、不依赖商业 App 的健身者。
TypeScript · ★ 10,991 · 🍴 1,308 · 📈 102 stars today
Agent workspace built on Cloudflare Workers for creating documents, building apps, and running agents with your company’s context and systems.
中文介绍 基于 Cloudflare Workers 构建的 Agent 工作空间，用于创建文档、开发应用和运行 Agent，并接入企业自身的上下文与内部系统。借助 Workers 的边缘运行和部署能力，适合企业在 Cloudflare 上搭建内部 AI 助手与自动化工作流的场景。
JavaScript · ★ 14,277 · 🍴 1,628 · 📈 111 stars today
Stremio - Freedom to Stream
中文介绍 Stremio 的网页版客户端，主打自由串流观影。依托插件化 addon 生态聚合影视资源与字幕，在浏览器里即可管理和播放媒体库，无需安装桌面客户端。适合已使用 Stremio 服务、希望在任意设备浏览器上直接观看的用户。
Shell · ★ 157,230 · 🍴 25,366 · 📈 687 stars today
A complete AI agency at your fingertips - From frontend wizards to Reddit community ninjas, from whimsy injectors to reality checkers. Each agent is a specialized expert with personality, processes, and proven deliverables.
中文介绍 一套开箱即用的 AI 代理团队集合：从前端实现、Reddit 社区运营，到创意注入和事实核查，每个 Agent 都是带有人格设定、工作流程与专长的角色。适合需要多角色协作完成内容、营销与开发任务的团队，直接组队调用即可分工干活。
C++ · ★ 1,317 · 🍴 120 · 📈 196 stars today
Pi-hole-class DNS ad-blocker on a $2 ESP32-C3 (no PSRAM): 537k domains as 40-bit FNV-1a hashes in flash, binary-searched. UDP DNS sinkhole + web dashboard. https://youtube.com/shorts/RaxszOUMi8E?feature=share
中文介绍 在仅售 2 美元的 ESP32-C3 上实现类 Pi-hole 的 DNS 广告拦截：把 53.7 万个域名以 40 位 FNV-1a 哈希存入 flash 并二分查找，无需 PSRAM，附带 UDP DNS sinkhole 与 Web 管理面板。适合想低成本、低功耗给全屋去广告的网络爱好者。
👍 234
A companion that talks with a person for months should come to understand them. It should remember what they said, infer who they are, and know when the past bears on the message in front of it. Testing this requires a real person's record, and such records are private, so benchmarks generate the pe
中文介绍 针对陪伴型 AI 需长期理解用户的能力评测，提出 RealCompanion 基准：以真实人的长期对话记录为基础（因隐私限制采用生成式替代），考察记忆、人物推断与上下文关联等推理能力，用可核验答案衡量模型对用户的理解程度。
👍 109
Large language models rely heavily on human text, which often conveys surface answers rather than the spatial and structural logic behind them. Protein folding is a natural testbed, because one solved structure yields thousands of exactly checkable spatial and topological statements. We ask: can lea
中文介绍 论文追问蛋白质折叠的学习能否泛化到更广泛的推理能力：以蛋白质折叠为测试床，一个解析出的结构可生成数千条可精确校验的空间与拓扑陈述，据此训练并评估 LLM 的结构化推理，检验其迁移效果。
👍 85
Vision-language-action (VLA) models have advanced robotic manipulation, but their zero-shot generalization in new tasks and environments remains limited, and their reliance on specialized training keeps them from benefiting directly from rapidly advancing general-purpose vision-language models (VLMs
中文介绍 针对 VLA 模型零样本泛化差、依赖专用机器人数据训练的问题，MotorMind 提出以通用 VLM 为骨架、补充运动相关模块的框架，使通用视觉语言模型可直接用于零样本机器人操作，提升新任务与新环境中的表现。
👍 69
The strong generalization performance of on-policy post-training paradigms has motivated studies of their parameter update behaviors. However, these studies treat the observed behaviors only as byproducts in on-policy training, overlooking their potential to serve as optimization principles for impr
中文介绍 论文研究 on-policy 后训练泛化能力强的成因，指出以往工作仅把参数更新行为视为副产品，转而将其作为优化原则，分析 on-policy 参数更新方向对泛化的决定性作用，并据此设计改进训练的方法。
👍 44
We introduce HyperBrowseComp, a multilingual and multimodal browsing benchmark comprising 423 manually authored and human-validated questions across 13 languages, written by native or highly proficient speakers. Questions are designed to be extremely challenging. Each question targets a concise, pub
中文介绍 提出 HyperBrowseComp，面向网页浏览智能体的多语言、多模态压力测试基准，含 423 道由母语者编写并人工校验的题目，覆盖 13 种语言，每题要求给出简短且可公开核验的答案，难度极高，用以暴露现有 agent 的短板。
👍 42
Masked diffusion language models (dLMs) offer a promising parallel alternative to autoregressive models for complex reasoning. However, they face a distinct credit-assignment challenge, since a few commitments during denoising sharply reduce the uncertainty over the remaining masked positions and sh
中文介绍 针对 masked diffusion 语言模型在去噪中少数位置确定后不确定性骤降、信用分配困难的问题，Pivot-SD 提出高效自蒸馏方法，聚焦关键决策位置施加监督，在复杂推理任务上提升生成质量并降低计算开销。
👍 42
Physical trajectories contain more than snapshots of a system: they also reveal how its states evolve under governing conditions. However, representation learning for parametric partial differential equations (PDEs) has largely relied on reconstruction-based objectives that emphasize recovering obse
中文介绍 针对参数化 PDE 的表示学习多依赖重建目标、只关注单帧观测的局限，PDE-JEPA 提出预测式表示学习框架，在隐空间建模状态随控制条件的演化动力学，从而学到更有利于下游预测的表征。
👍 39
Legal language models require reward signals that capture not only answer correctness but also the multidimensional quality of legal responses. Existing reward methods, however, often rely on coarse-grained holistic judgments, providing limited domain specificity and interpretability. We introduce L
中文介绍 针对法律语言模型奖励信号粗糙、缺乏领域特异性与可解释性的问题，LexReward 提出基于分类体系（taxonomy）的奖励框架，从答案正确性与多维法律回答质量两方面评分，为法律领域后训练提供更细粒度的奖励。
👍 39
AI-generated content, often called AI slop, is increasingly common everywhere, particularly in academia. Slop in AI-generated scientific papers, however, has more complex patterns that cannot be easily detected by existing token-based AI detectors. Each part of such a paper looks plausible while the
中文介绍 针对 AI 生成科学论文中难以被现有 token 级检测器识别的「学术垃圾」，论文构建相应基准并刻画其模式：各部分看似合理但整体缺乏实质，同时给出检测与缓解方法，用于评估和改进 AI 生成论文的质量。
👍 38
On-policy distillation (OPD) trains a student on the responses it generates. Existing LLM multi-teacher OPD transfers what specialists predict through their output distributions. We introduce Latent-MOPD, to our knowledge the first representation-level multi-teacher OPD method for LLMs. It integrate
中文介绍 现有多教师 on-policy 蒸馏仅在输出分布层面迁移专家知识，Latent-MOPD 提出首个面向 LLM 的表征级多教师 on-policy 蒸馏方法，将多个专家模型的隐层表征融合进学生训练，以整合各自专长。
👍 37
Existing Simulink benchmarks mainly evaluate whether generated models compile, execute, or resemble a reference model. These criteria do not establish whether a model satisfies its engineering requirements. We introduce SimuVerity, a benchmark of 101 text-to-executable Simulink model-generation task
中文介绍 现有 Simulink 基准只看模型能否编译、运行或与参考模型相似，无法判断是否满足工程需求。SimuVerity 构建 101 个文本到可执行 Simulink 模型的生成任务，按工程需求而非表面相似度评估智能体的建模能力。
👍 36
As LLM agents undertake increasingly complex, long-horizon tasks, verifying their outputs becomes increasingly challenging. We study how verification capability can be strengthened with a fixed base model, without access to reference answers or grading rubrics at test time. Repeated sampling yields
中文介绍 针对长程任务中 LLM agent 输出难以验证、测试时又无参考答案与评分标准的问题，VeriHarness 研究如何用固定基座模型强化验证能力，通过重复采样与 agent 化验证流程提升验证的规模与可靠性。
👍 34
Long-video generation and world models have shown strong potential for interactive entertainment and embodied simulation by predicting future observations conditioned on user actions and historical memory. However, as memory sequences grow longer and their structures become increasingly complex, man
中文介绍 针对长视频生成与世界模型中记忆序列不断增长、结构日益复杂而难以管理的问题，该工作提出 Spatial Memory Intelligence，为世界模型赋予理解驱动的长期空间记忆，使其在用户动作条件下生成更一致的未来观测。
👍 33
We understand little about how capabilities acquired in one language carry over to another, or what governs this transfer: evaluations rely on incomparable, saturation-prone datasets and rarely examine its determinants jointly. Identifying what predicts transfer would let us avoid exhaustive evaluat
中文介绍 论文构建多语言 GSM-Symbolic 数据集，系统考察某一语言中获得的能力如何迁移到其他语言，并联合分析影响迁移的决定因素，避免对每种语言逐一穷举评测，从而揭示跨语言能力迁移的关键变量。
👍 30
Scientific progress emerges from a longitudinal ecosystem in which researchers, institutions, funding agencies, collaboration networks, and the scientific literature co-evolve. As AI becomes increasingly involved throughout the scientific research cycle, understanding these interconnected and evolvi
中文介绍 将科研视为研究者、机构、资助方、合作网络与文献共同演化的长期生态系统，论文提出基于 LLM 的闭环模拟框架，用于研究 AI 深度参与科研流程后，各要素相互作用对科学进展的影响。
👍 20
Autoregressive (AR) video models excel at causal generation, but their reliance on next-chunk prediction confines them to a short-sighted, reactive paradigm. This limitation is particularly consequential for reasoning-oriented generation, where achieving a target outcome through valid intermediate s
中文介绍 自回归视频模型依赖下一片段预测，只能短期、反应式生成，难以通过有效的中间步骤达成目标。ProAR 提出前瞻式推理方法，让 AR 视频模型能够规划并生成通向目标结果的中间过程，提升推理导向的生成能力。
👍 18
We explore catastrophic forgetting in the context of large pre-trained models. By considering forgetting as a geometric problem in the input space of each weight matrix, we uncover a natural retention objective under which updates produced by gradient-based optimizers are suboptimal. Following this
中文介绍 论文将大预训练模型的灾难性遗忘视为每个权重矩阵输入空间中的几何问题，由此导出一个自然的保留目标，并指出梯度优化器产生的更新在该目标下并非最优，据此提出 Local Support Learning 缓解遗忘。
👍 17
Policies with similar mean returns can differ sharply in rare failures, yet estimating lower-tail conditional value-at-risk (CVaR) accurately can require many costly rollouts. When different conditional components of a stochastic workflow can be queried separately, we ask how to allocate a fixed eva
中文介绍 平均回报相近的策略在罕见失败上可能差异巨大，而准确估计下尾 CVaR 往往需要大量昂贵采样。论文提出 Tail-Influence Sampling，在可分别查询随机流程各条件分量时，将固定评估预算分配给对尾部风险影响最大的分量。
👍 17
Modern chess engines are silent experts: they play at a superhuman level, but do not offer explanations for their play. On the other hand, language models (LMs) can generate plausible-sounding explanations, but their weak playing strength limits the utility of their explanations. We introduce Queen,
中文介绍 现有国际象棋引擎棋力超人类却从不解释着法，而语言模型能生成看似合理的解释但棋力偏弱。论文提出 Queen，使语言模型既下出较强着法又能说明理由，兼顾棋力与可解释性。
👍 17
LLM-guided evolutionary methods, such as AlphaEvolve, have emerged as powerful approaches for challenging computational optimization problems, such as circle packing. However, prior work typically optimizes performance gain over a fixed number of iterations. We argue that practical optimization shou
中文介绍 针对 AlphaEvolve 等 LLM 引导演化方法只优化固定迭代次数下的性能、忽视算力成本的问题，FrugalEvo 提出成本感知的程序演化框架，在圆堆积等优化任务中显式权衡性能增益与调用开销。
@BrianRoemmele · 489.9K 粉丝 · 1.1M 阅 · 533 赞 · 85 转
You do not build a Grok Bot. You hire one. The difference is the whole article. A chatbot answers a question and forgets the room. A Grok Bot has a name, a job, a conversation that persists, and a
中文介绍 Brian Roemmele 提出「雇佣」而非「构建」Grok Bot 的思路：聊天机器人答完就忘，而 Grok Bot 有名字、职位和持久对话记忆，是首个零人类公司的用工经验分享。
@stevenmarkryan · 458.8K 粉丝 · 124.9K 阅 · 541 赞 · 84 转
Incredible things are happening in AI. What’s more incredible is that it’s still EXTREMELY early. As in, Day 0.000. Things are only going to get crazier from here. In case you’ve been living under a
中文介绍 Steven Mark Ryan 谈算力即权力：AI 领域正在发生惊人变化，但整体仍处在 Day 0.000 的极早期阶段，接下来的发展只会更疯狂。属于长线叙事视角的观点帖。
@TryLiveAvatar · 1.1K 粉丝 · 123.4K 阅 · 692 赞 · 188 转
The best way to learn a language was always a patient tutor. For the first time, everyone can have one. Here's what that changes for learners, teachers and schools. The hardest part is speaking Ask
中文介绍 LiveAvatar 推出面向语言学习者的 AI 私教：学语言最难的是开口说，而耐心的口语辅导此前稀缺。帖子讨论人人都有 AI 陪练后，对学习者、老师和学校意味着什么。
@gregisenberg · 724.2K 粉丝 · 44.1K 阅 · 569 赞 · 53 转
For the FIRST TIME in HISTORY, you can describe a physical product in one sentence and, a week later, hold it in your hands, cut from metal and built to your exact measurements. That's VIBE
中文介绍 Greg Isenberg 提出「氛围制造」概念：史上第一次可以用一句话描述物理产品，一周后就拿到按精确尺寸用金属加工出来的实物，分享 AI 驱动小批量硬件制造的新玩法。
@sermakarevich · 549 粉丝 · 33.0K 阅 · 520 赞 · 64 转
An article for everyone who ships, buys, or signs off on software built on large language models (LLMs): engineers, product managers, and the CEO. Written in plain language, from the big picture down
中文介绍 面向 LLM 应用落地的 evals 科普长文，读者覆盖工程师、产品经理到 CEO，用平实语言从整体图景讲到具体做法，解决「怎么判断一个 AI 系统是否真的可用」这一签收前的核心问题。
@onlyonealexia · 1.6K 粉丝 · 26.5K 阅 · 509 赞 · 53 转
A List of Web3, AI & Web2 Hackathons Compiled by @onlyonealexia There are a lot of hackathon lists that look impressive until you actually open the links. I compiled a list of ongoing and upcoming
中文介绍 整理了一份 Web3、AI 与 Web2 黑客松清单：作者吐槽很多清单看着唬人、点开链接才发现没用，于是自己汇总了正在进行和即将开始的活动，供参赛者直接筛用。
@rehan_shei · 11.9K 粉丝 · 24.9K 阅 · 520 赞 · 49 转
A week ago I pointed Opus 5.5 at Terraria before bed. I woke up to a fully working, reverse engineered version of the game. Then I added a homing missile launcher, a tactical nuke and a bunch of new
中文介绍 博主让 Opus 5.5 在睡前跑 Terraria 逆向工程，醒来就得到一份可运行的完整复刻版，还顺手加了追踪导弹、战术核弹等新武器，用来论证「所有游戏现在都算开源」。
@goyal__pramod · 11.4K 粉丝 · 24.7K 阅 · 504 赞 · 59 转
Now I am aware you must have been looking forward to the second part of my CUDA blog (if you haven't read it yet, check it out!), but hey! A man can have varied interests. And I believe if you are
中文介绍 从零实现强化学习系列第一篇。作者暂时搁置 CUDA 博客的第二部分，转向自己的另一个兴趣方向，开启 RL 从零手写的新连载，适合想补底层实现的读者跟进。
For all the data that AI systems continually amass and analyze, enterprise AI agents often suffer from a curious shortcoming: a lack of knowledge. More than data, knowledge is the understanding of what the data means in the context of individual organizations. AI agents need this understanding to re
中文介绍 企业AI智能体在持续积累和分析数据的同时，常存在一个短板：缺乏知识。知识不只是数据，而是理解数据在具体组织语境中的含义。文章探讨如何把AI智能体与企业知识连接起来。
How OpenAI is approaching text watermarking under EU rules. Learn where watermarks apply, how detection works, and why access starts with researchers.
中文介绍 OpenAI介绍其在欧盟文本溯源规则下推进文本水印的做法，涵盖水印的适用范围、检测机制的运作方式，以及为何相关访问权限首先面向研究人员开放。
In 2026, the question for enterprise AI is no longer whether predictive models can outperform statistical forecasts—that argument is settled. The big question now is how to enable predictive systems to act on their own conclusions without drifting from business intent. The frontier has moved from pr
中文介绍 到2026年，企业AI的争论已不再是预测模型能否胜过统计预测，而是如何让预测系统依据自身结论采取行动，同时不偏离业务意图。
OpenAI introduces a new visual ad format in ChatGPT and expands measurement tools, attribution partnerships, and brand suitability for advertisers.
中文介绍 OpenAI在ChatGPT中推出新的视觉广告形式，并扩展面向广告主的衡量工具、归因合作关系与品牌适配能力。
Over the summer I talked to the CEO of Springboards, a startup building an LLM that’s designed to come up with a wider variety of responses than its mainstream rivals do. At the start of the call, he said something that’s been stuck in my head since: “We often say that we’re a self-loathing AI…
中文介绍 作者回顾与初创公司Springboards首席执行官的交谈，该公司打造的LLM旨在给出比主流对手更多样的回答；文章借此讨论人们一面反感AI、一面又离不开它的矛盾。
Yossi Matias, Vice President & Head of Google Research, explores how AI is beginning to reshape biology, infrastructure, manufacturing, and science, and why its greatest impact may come when it intersects with other fields. Step inside the newsroom with our MIT Technology Review editors for sharp an
中文介绍 谷歌研究副总裁兼负责人Yossi Matias在EmTech Future 2026上探讨AI如何开始重塑生物学、基础设施、制造业与科学，并认为其最大影响可能来自与其他领域的交叉。
中文介绍 该期TLDR AI简报的要点涉及Prime Inference、智能体数量增长以及Claude academy（Claude学院）三个话题。
中文介绍 Hugging Face博客文章，标题为「智能体说它做完了，数据库却不同意」，主题涉及AI智能体与数据库实际状态不一致的问题。
a quiet day.
中文介绍 Latent Space的AINews栏目称当日行业消息不多，是相对平静的一天。
Learn how startups can choose GPT-6 models, tune reasoning effort, improve prompts and skills, coordinate tools, and prepare workflows for production.
中文介绍 OpenAI发布GPT-6系列模型使用指南，介绍初创公司如何选择GPT-6模型、调整推理强度、优化提示与技能、协调工具，并为生产环境准备工作流。
Enterprise AI is no longer a future ambition. It is in full operational flight. Model capabilities are advancing faster than most organizations can absorb, while the cost of performance continues to fall. Globally, AI investment is set to reach $2.5 trillion in 2026, up 44% from the previous year. F
中文介绍 企业AI已进入实际运营阶段。2026年全球AI投资预计达2.5万亿美元，同比增长44%；模型能力提升速度快于多数组织的吸收能力，而性能成本持续下降。
中文介绍 Allen AI开源AstaBrief，这是其Asta产品中用于快速生成报告的模型。
After leading Meta’s Llama models, Ahmad Al-Dahle is now transforming Airbnb with AI — from how its teams develop products to how it serves guests.
中文介绍 在领导Meta的Llama模型之后，Ahmad Al-Dahle现正以AI改造Airbnb，涵盖团队的产品开发方式以及面向房客的服务体验。
On an afternoon in Seoul in March 2016, I watched a program I helped build put a stone on the fifth line of a Go board in what looked like a gift to its human opponent. Move 37 in game two of the five-game match looked so absurd that some commentators thought it was a…
中文介绍 作者回忆2016年在首尔观看的一场围棋比赛中看似荒谬的第37手，以此论证不应被表象迷惑：大语言模型并不具备推理能力。
the minimalist harness goes stable... and TypeScript!
中文介绍 极简框架Pi发布1.0稳定版，并加入TypeScript支持；内容还涉及Pi Durable与AIE NYC。
美股方面，SPY 报 774.83、单日 +0.67%，MACD 金叉（1.4745 对 1.0094），距 52 周高 -0.58%；QQQ 报 756.20、单日 +0.88%、近 5 日 +2.67%，RSI14 68.4，距 52 周高仅 -0.09%，指数层面偏多。个股存在分化：MSFT 单日 +1.48%、NVDA +2.12% 且后者距 52 周高 -0.50%，而 AAPL 单日 -0.24%、近 5 日 -1.63% 并现 MACD 死叉，META 同为 MACD 死叉，TSLA 趋势读数中性。加密市场恐慌贪婪指数 70（贪婪），总市值 2.94T USD、24 小时 -2.23%，BTC 主导率 58.7%、ETH 11.3%；BTC 85921.22 微跌 0.65%，仍站上 SMA20/50/200 呈多头排列，RSI14 65.5，但 MACD 2005.0 已略低于信号 2063.2；ETH RSI 62.2、SOL RSI 64.3 亦为多头排列，SOL MACD 出现死叉。中概技术面整体偏弱：BABA 单日 +4.68% 并现 MACD 金叉却仍空头排列，PDD RSI 43.8、JD RSI 39.2，腾讯 RSI 41.8、MACD 死叉且接近 52 周低。商品外汇方面，黄金 RSI 34.6、MACD -66.56 低于信号 -44.21 且空头排列；WTI 原油单日 -2.04%、近 5 日 -3.62% 并跌破 SMA20 95.93；美元指数 RSI 73.2 与 10Y 收益率 RSI 76.2 双双超买，美元兑人民币 RSI 38.9 且接近 52 周低；VIX 15.52 处于 52 周低位区域。
现价 756.20，单日 +0.88%、近 5 日 +2.67%，距 52 周高仅 -0.09%；RSI14 68.4 逼近超买区但未破 70，MACD 9.1262 高于信号线 7.0153 且柱状为正，价格位于 SMA20 729.64、SMA50 718.26 与 SMA200 669.33 之上，呈多头排列，动量读数整体偏强。
收益率报 5.31，单日 +0.64%、近 5 日 +1.35%，距 52 周高仅 -0.71%；RSI14 76.2 已进入超买区，MACD 0.1367 高于信号 0.1203 且柱状为正，价格同时站上 SMA20 5.07、SMA50 4.84 与 SMA200 4.46，多头排列，上行动量读数在候选资产中最为突出。
现价 346.47，单日 +0.86%、近 5 日 +1.09%，但仍低于 52 周高 15.21%；RSI14 52.4 居中，MACD -0.2233 刚上穿信号 -0.3624 形成金叉，然柱状仍处零轴下方，价格在 SMA20 343.02 与 SMA50 344.90 之间窄幅震荡，动能与趋势读数均未给出明确方向。
腾讯报 423，近 5 日 -3.11%，距 52 周高 -38.02% 并接近 52 周低；RSI14 41.8 处于偏弱区间，MACD -4.5082 低于信号 -4.0645 呈死叉，价格同时低于 SMA20 432.71、SMA50 447.28 与 SMA200 497.98，空头排列，技术状态偏下行。
黄金期货报 4165.7，单日 +0.08%、近 5 日 -0.06%，距 52 周高 -25.43%；RSI14 34.6 已接近超卖，MACD -66.5575 低于信号 -44.21 且柱状为负，价格低于 SMA20 4318.78、SMA50 4368.17 与 SMA200 4552.53，呈空头排列，下行动量占优。
VIX 恐慌指数
10Y 美债收益率 (%)
美元指数 DXY
S&P 500 ETF
Nasdaq 100 ETF
Apple
Microsoft
Nvidia
Alphabet
Tesla
Meta
Bitcoin
Ethereum
Solana
阿里巴巴 (BABA)
拼多多 (PDD)
京东 (JD)
腾讯控股 (0700.HK)
黄金期货
WTI 原油期货
美元 / 人民币
本报告仅基于公开行情数据计算出的技术指标读数，过去走势不代表未来表现，任何金叉、死叉或超买超卖信号均可能失效；内容仅供技术指标解读参考，不构成任何投资建议，请读者自行判断并承担风险。
With neither candidate getting more than 50% of the vote, the election will go to a run-off on 25 October.
中文摘要 巴西右翼候选人弗拉维奥·博索纳罗赢得总统选举首轮投票，因无候选人得票超过50%，选举将于10月25日进入第二轮决选。
What happened in Mangala, where residents were left with little outside support, helps explain how this outbreak became the worst seen in the Democratic Republic of Congo.
中文摘要 刚果民主共和国的埃博拉疫情迅速恶化，曼加拉地区居民几乎未获外部支持，此次疫情已成为该国迄今最严重的一次暴发。
Sources say the opposition’s target will be below Labor’s but won’t match One Nation’s goal to limit net arrivals to 130,000 Get our new political email, free app or daily news podcast The Coalition is expected to reveal deep cuts to immigration today which the party says would maintain working holi
中文摘要 澳大利亚反对党联盟预计将公布大幅削减海外移民的方案，消息人士称其目标低于工党，但不会达到「单一民族党」把净入境人数限制在13万人的水平。
Hispanic voters are twice as likely to favour Democratic Party over Trump's Republican Party in the November elections.
中文摘要 中期选举前，特朗普在拉美裔选民中的支持率跌至新低。民调显示，在11月选举中拉美裔选民支持民主党的可能性是支持共和党的两倍。
In Iraq, NPR attends trials of French ISIS suspects after transfer from Syrian camps.
中文摘要 伊拉克开庭审理此前关押在叙利亚的极端组织「伊斯兰国」成员，NPR记者旁听了法国籍嫌犯的庭审。
Student groups call for the abolition of Greek life and more punitive action in the wake of fraternity assault allegations.
中文摘要 康奈尔大学学生团体就兄弟会轮奸指控发起抗议，在校园建筑上涂鸦，要求取缔希腊式社团生活并采取更严厉的惩处措施。
The Pentagon confirmed on Sunday it had removed the B1 bombers from the base in England back to their home stations in America.
中文摘要 特朗普称由于存在「威胁」，美国将B1轰炸机从英国皇家空军费尔福德基地撤回本土。五角大楼已于周日证实这批轰炸机返回美国驻地。
Flávio Bolsonaro’s first round lead leaves President Luiz Inácio Lula da Silva three weeks to find the votes to stop Brazil’s, and Latin America’s, rightward shift.
中文摘要 弗拉维奥·博索纳罗在巴西总统选举首轮投票中领先，令右翼大受提振；总统卢拉仅剩三周时间争取选票，以阻止巴西乃至拉美整体右转的趋势。
Donald Trump says the US was looking "very strongly" at the case of a lab technician who died at a research centre in Siberia.
中文摘要 俄罗斯西伯利亚一家研究中心的实验室技术员死亡后，特朗普表示美方正在「非常密切地」关注此案。
The country’s consumer watchdog has said only that “an employee of an anti-plague institute in the Irkutsk Region” contracted pneumonia.
中文摘要 俄罗斯消费者权益监管部门试图平息西伯利亚暴发瘟疫的传言，仅表示伊尔库茨克州一家抗鼠疫研究所的一名员工感染了肺炎。
The offensive is an effort to wrest control of the Bab al-Mandab Strait from the Houthis, a militia in Yemen backed by Iran.
中文摘要 沙特与也门部队对胡塞武装发动进攻，试图从这支受伊朗支持的民兵组织手中夺回曼德海峡及红海沿岸的控制权。
The crash was one of Nigeria’s worst military aviation accidents in years. The country has a history of fatal military aircraft crashes.
中文摘要 尼日利亚一架军用直升机坠毁，机上载有32人，这是该国多年来最严重的军用航空事故之一。尼日利亚过去曾多次发生军机致命坠毁事件。
France prepares for a day of protests in support of students who’ve been demanding more investment in education.
中文摘要 法国准备迎接全国性的校园抗议日，以支持要求增加教育投入的学生。此前的抗议活动中已出现人员受伤和大规模逮捕。
Michael Olise scores twice and assists as France come from 1-0 with 13 minutes left to beat Belgium 4-1
中文摘要 法国队在比赛还剩13分钟时以0比1落后，替补登场的迈克尔·奥利斯梅开二度并送出助攻，最终法国以4比1大胜比利时。
Neighbouring countries say coastal states must govern Red Sea as fighting escalates in northern Ethiopia.
中文摘要 埃塞俄比亚总理坚持要求获得红海出海通道，尽管北部战事升级带来地区冲突风险。邻国表示，红海应由沿岸国家共同管理。
Asian stocks were set to open higher after a tech-led rally pushed the S&P 500 Index near a record as investors largely shrugged off concerns about bond yields at multi-decade highs. US oil prices extended declines.
中文摘要 科技股领涨推动标普500指数逼近历史高点，投资者基本无视债券收益率处于数十年来高位的担忧。亚洲股市料将高开，美国油价延续跌势。
The world's fastest growing major economy has one of the worst performing major equity markets in 2026.
中文摘要 印度是全球增长最快的主要经济体，但其股市却是2026年表现最差的主要股票市场之一。文章列举了五个原因解释这一背离现象。
A growing number of parents are opening retirement funds for their children.
中文摘要 越来越多父母为子女开设退休基金。一对父母每月为幼儿和婴儿存入100英镑养老金，用于长期储蓄。
Malaysia’s consumer and construction stocks may emerge as beneficiaries of next year’s spending plan as Prime Minister Anwar Ibrahim seeks to ease living costs while maintaining fiscal prudence.
中文摘要 马来西亚总理安瓦尔·易卜拉欣在设法缓解生活成本压力的同时保持财政审慎，明年的支出计划或使消费类与建筑类股成为受益者。
Asian stocks were set to open higher after a tech-led rally pushed the S&P 500 index near a record as investors largely shrugged off concerns about bond yields at multi-decade highs. US oil prices extended declines.
中文摘要 科技股领涨推动标普500指数逼近历史高点，投资者基本无视债券收益率处于数十年来高位的担忧。亚洲股市料将高开，美国油价延续跌势。
Oil held a two-day drop, as rising Persian Gulf exports and a steep price cut by Saudi Arabia reinforced signs of a loosening market.
中文摘要 油价延续两日跌势。波斯湾出口增加，加之沙特大幅下调价格，进一步显示市场供应趋于宽松。
The growing demand for capital to fund artificial intelligence investment is filtering into a little-watched corner of the asset-backed securities market: bonds secured by equipment loans and leases.
中文摘要 人工智能投资所需资金需求上升，正渗入资产支持证券市场中较少受关注的一角，即由设备贷款和租赁担保的债券。Stonebriar与Wingspire达成相关交易。
Even at the time, it was considered an unconventional alliance: Mark Walter committing $160 million through his education-focused nonprofit to help a Chicago credit investment firm comply with Dodd-Frank regulations in return for funding.
中文摘要 马克·沃尔特通过其关注教育的非营利组织承诺出资1.6亿美元，帮助芝加哥一家信贷投资公司满足《多德-弗兰克法案》监管要求，以换取资金支持，该联盟当时即被视为非常规安排。
Babylist CEO Jennifer Hyman discusses the company’s new 20,000-square-foot showroom in Manhattan’s SoHo neighborhood, which opened at the end of September and is its largest retail location yet. Hyman says the space is designed to help expecting and new parents research major purchases with expert g
中文摘要 Babylist首席执行官珍妮弗·海曼介绍公司在曼哈顿SoHo区新开设的2万平方英尺展厅，该店于9月底开业，是其最大零售门店，旨在帮助准父母和新手父母了解重要产品。
Paramount Skydance Corp. outlined the executive leadership team that will helm the new company resulting from its acquisition of Warner Bros. Discovery Inc. David Ellison will continue to serve as chief executive officer and chairman, overseeing the company’s long-term strategy and creative vision,
中文摘要 派拉蒙Skydance公司公布收购华纳兄弟探索公司后新公司的管理层团队。大卫·埃里森将继续担任首席执行官兼董事长，负责公司长期战略。
A rally in giant technology companies drove stocks higher in the Monday session, overshadowing concerns about still-elevated oil prices and bond yields. The S&P 500 rose to a striking distance of a record while the Nasdaq 100 reached a fresh peak. Political upheaval in Europe and fiscal worries drov
中文摘要 大型科技股上涨推动周一美股走高，标普500指数逼近历史纪录，纳斯达克100指数创下新高，市场对高企的油价与债券收益率的担忧被盖过。
British Airways will introduce what it called the world’s biggest business class on its upgraded double-decker Airbus SE A380 jets, targeting premium passengers as airlines rush to cash in on upscale travel.
中文摘要 英国航空将在升级后的空客A380双层客机上推出其称为全球最大的商务舱，瞄准高端旅客，因各航空公司竞相从高端旅行需求中获利。
42 回复 · 程序员 节点
9 回复 · 程序员 节点
10 回复 · 程序员 节点
6 回复 · 程序员 节点
27 回复 · 程序员 节点
7 回复 · 程序员 节点
15 回复 · Apple 节点
26 回复 · 程序员 节点
18 回复 · Apple 节点
35 回复 · Apple 节点
该源今日无内容。