Python · ★ 69,572 · 🍴 10,023 · 📈 3,563 stars today
利用AI大模型，一键生成高清短视频 Generate short videos with one click using AI LLM.
中文介绍 基于 AI 大语言模型的一键短视频生成工具，用户只需输入主题关键词，系统自动完成文案撰写、素材匹配、语音合成和字幕生成，输出高清短视频。适合内容创作者、自媒体运营者快速批量产出视频，大幅降低视频制作门槛，支持多种 AI 模型接入。
Python · ★ 129,878 · 🍴 8,914 · 📈 1,876 stars today
Python tool for converting files and office documents to Markdown.
中文介绍 微软开源的 Python 文档转换工具，可将 Word、Excel、PowerPoint 等 Office 文档及多种常见文件格式批量转换为 Markdown 格式。适合需要将文档内容导入知识库、RAG 管道或 LLM 处理流程的开发者，简化文档预处理步骤。
TypeScript · ★ 18,116 · 🍴 1,376 · 📈 354 stars today
Official Compound Engineering plugin for Claude Code, Codex, Cursor, and more
中文介绍 Every Inc. 发布的官方 Compound Engineering 插件，支持 Claude Code、OpenAI Codex、Cursor 等主流 AI 编程工具。通过统一插件规范，帮助开发者在不同 AI 编码环境中复用工程实践方法和工作流，提升跨工具协作效率。
TypeScript · ★ 48,392 · 🍴 6,861 · 📈 575 stars today
The open alternative to Salesforce, designed for AI.
中文介绍 面向 AI 时代设计的开源 CRM 系统，定位为 Salesforce 的开放替代方案。提供客户管理、销售流程跟踪等核心功能，并在架构上为 AI 集成预留接口。适合希望掌控数据主权、避免高昂 SaaS 费用的企业或开发团队自托管部署。
Python · ★ 127,847 · 🍴 20,907 · 📈 460 stars today
Claude Code is an agentic coding tool that lives in your terminal, understands your codebase, and helps you code faster by executing routine tasks, explaining complex code, and handling git workflows - all through natural language commands.
中文介绍 Anthropic 官方发布的终端原生 AI 编程助手，深度理解整个代码库上下文，可自动执行日常编码任务、解释复杂逻辑、管理 Git 工作流。以 Agent 模式运行，开发者无需切换界面即可在命令行中完成代码审查、重构、提交等操作，显著提升开发效率。
Shell · ★ 28,074 · 🍴 2,080 · 📈 2,066 stars today
Taste-Skill - gives your AI good taste. stops the AI from generating boring, generic slop
中文介绍 针对 AI 生成内容同质化问题的提示词技能文件，通过注入「审美偏好」约束，引导 AI 模型产出更具个性、避免模板化套话的文本内容。适合写作者、内容创作者在使用 Claude 等 LLM 时提升输出质量，抑制 AI 惯用的平庸表达模式。
TypeScript · ★ 1,287 · 🍴 110 · 📈 129 stars today
Cursor plugin specification and official plugins
中文介绍 Cursor 编辑器官方插件规范及官方插件仓库，定义了第三方插件与 Cursor AI 编程环境的集成标准和接口协议。供插件开发者参考规范构建扩展，也包含官方维护的示例插件，是生态建设的核心基础设施。
Rust · ★ 7,249 · 🍴 442 · 📈 680 stars today
A fast, helpful, and open-source document parser
中文介绍 LlamaIndex 团队推出的轻量级开源文档解析器，专注于快速、准确地解析 PDF、HTML 等多种格式文档。相比重型解析方案更轻便，适合在 RAG 应用、文档问答系统中作为前置解析模块，降低构建 LLM 知识管道的复杂度。
Python · ★ 1,234 · 🍴 148 · 📈 346 stars today
A platform for reproducible world model research and evaluation
中文介绍 面向世界模型研究的可复现实验平台，提供标准化的训练、评估流程和基准测试套件。帮助 AI 研究者在统一环境下对比不同世界模型方法的性能，解决该领域实验结果难以复现和横向比较的痛点，适合学术界和工业界的世界模型研究团队使用。
★ 49,619 · 🍴 5,196 · 📈 1,564 stars today
An advanced guide to learn English which might benefit you a lot 🎉 . 离谱的英语学习指南/英语学习教程/英语学习/学英语
中文介绍 系统性的英语自学进阶指南，涵盖听说读写各项技能的学习方法论、资源推荐和训练路径规划。内容基于实践经验整理，适合有一定基础、希望突破瓶颈的中国学习者，尤其关注如何高效利用碎片时间和互联网资源实现英语水平的实质性提升。
Jupyter Notebook · ★ 2,561 · 🍴 313 · 📈 64 stars today
中文介绍 该仓库名为 ESM，归属于 Biohub 组织，与 Meta 的蛋白质语言模型 ESM（Evolutionary Scale Modeling）系列相关，用于蛋白质序列表示学习与功能预测。适合生物信息学研究者和计算生物学团队用于蛋白质结构预测和序列分析任务。
TypeScript · ★ 26,941 · 🍴 2,656 · 📈 294 stars today
Project N.O.M.A.D, is a self-contained, offline survival computer packed with critical tools, knowledge, and AI to keep you informed and empowered—anytime, anywhere.
中文介绍 面向极端离网场景设计的自包含离线生存计算机系统，内置关键工具、生存知识库和本地 AI 推理能力，无需网络连接即可运行。适合野外探险、灾难应急或基础设施崩溃场景下使用，为用户在断网环境中提供信息查询和 AI 辅助决策能力。
HTML · ★ 171,737 · 🍴 3,340 · 📈 1,317 stars today
DigitalPlat FreeDomain: Free Domain For Everyone
中文介绍 DigitalPlat 提供的免费域名注册服务项目，让任何人都能免费获取域名。适合个人开发者、开源项目或学生搭建网站时降低初始成本，通过该平台可申请免费子域名或特定后缀域名，降低互联网入门门槛。
JavaScript · ★ 198,550 · 🍴 30,507 · 📈 1,413 stars today
The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.
中文介绍 面向 AI 编程 Agent 的性能优化增强框架，为 Claude Code、Codex、Cursor 等工具提供技能扩展、本能行为、持久记忆和安全加固模块。通过标准化的「Agent 马具」接口，帮助开发者在多种 AI 编程工具中复用最佳实践，提升 Agent 的自主编程能力和可靠性。
★ 6,966 · 🍴 491 · 📈 618 stars today
A skill file for removing AI tells from prose
中文介绍 专为消除 AI 写作痕迹设计的技能提示词文件，通过一套规则约束引导 LLM 规避「AI 腔」常见模式，如过度使用转折词、格式模板化、表达空洞化等问题。适合写作者、编辑在润色 AI 生成内容时使用，让文章读起来更接近人类自然表达。
Jupyter Notebook · ★ 41,550 · 🍴 8,271 · 📈 180 stars today
Data Engineering Zoomcamp is a free 9-week course on building production-ready data pipelines. The next cohort starts in January 2026. Join the course here 👇🏼
中文介绍 DataTalks.Club 出品的免费 9 周数据工程训练营，系统讲授构建生产级数据管道所需的全套技术栈，涵盖批处理、流处理、数据仓库、编排工具等核心模块。下一期将于 2026 年 1 月开课，适合希望转型或深入数据工程领域的开发者和数据从业者。
Markdown · ★ 507,331 · 🍴 48,167 · 📈 867 stars today
Master programming by recreating your favorite technologies from scratch.
中文介绍 通过从零复现经典技术来深入掌握编程原理的学习资源合集，涵盖自制 Git、数据库、编译器、Web 服务器、操作系统等数十个方向的教程链接。适合有一定编程基础、希望突破「只会调用 API」阶段、真正理解底层原理的开发者系统性进阶学习。
👍 12
Vision-language models (VLMs) achieve strong performance on spatial reasoning benchmarks, yet it remains unclear whether this reflects structured 3D understanding or reliance on statistical shortcuts in natural images. We introduce a representation-level analysis framework that constructs minimal co
👍 2
Robot manipulation critically depends on perception that preserves the action-relevant aspects of a scene. Yet most robot learning pipelines are built upon visual encoders pre-trained for static recognition or vision-language alignment, leaving motion understanding to downstream policies. We introdu
👍 0
Recent advances in Vision-Language Models (VLMs) have achieved impressive performance across many tasks, yet prior studies report unsatisfactory performance when applying large language or multimodal models to finding abnormal patterns in sequential data. Public anomaly detection benchmarks typicall
👍 0
Large language models (LLMs) exhibit systematic political bias across a variety of sensitive contexts. We find that LLMs handle counterpart topics from opposing political sides asymmetrically. We refer to this phenomenon as covert political bias and identify 7 categories of techniques through which
👍 3
One-shot Program-of-Thought (PoT) emits a Python program that prints a primitive-action plan; a single invalid action silently invalidates the trajectory. We introduce RePoT (Recoverable PoT): a deterministic verified replay that walks the plan through the environment to its first invalid transition
👍 15
Explaining why dense retrievers assign high relevance scores remains challenging because retrieval decisions are made through opaque high-dimensional embeddings. Existing explanations often focus on surface signals, such as lexical matches, token alignments, or post-hoc textual rationales, and thus
👍 1
Tool retrieval over large API catalogs is a core bottleneck for LLM agents: user queries arrive in colloquial, often underspecified language, while the catalog uses technical API vocabulary that no fixed encoder can bridge on its own. The two dominant training approaches, contrastive encoder fine-tu
👍 23
Video large language models (Video-LLMs) have demonstrated strong capabilities in video understanding tasks. However, their practical deployment is still hindered by the inefficiency introduced by processing massive amounts of visual tokens. Although recent approaches achieve extremely low token ret
👍 7
The design space of agentic AI inference spans two extremes: frontier large language models (LLMs), typically hosted in the cloud and offering strong performance across a wide range of tasks at substantially high cost, and more cost-efficient small language models (SLMs), which are amenable to on-de
👍 3
Natural generation allows Large Language Models (LLMs) to produce free-form responses with rich reasoning, yet the lack of structure makes outputs difficult to verify. Conversely, constrained decoding ensures standardized formats but can inadvertently restrict reasoning capabilities by imposing cons
👍 6
We address the task of generating physically accurate and visually faithful 4D Human-Object Interaction (HOI). Given a static 3D human and target object represented as 3D Gaussian Splats (3DGS), our goal is to synthesize dynamic scenes where the human actively engages with the object through actions
👍 19
Activation-based control steers large language models (LLMs) by intervening on their internal representations during inference, and has emerged as an effective paradigm for controlling behaviors such as persona and style. However, existing methods often rely on fixed steering directions or task-spec
👍 1
We study two-level autoresearch for cooperation: an outer-loop AI agent autonomously redesigns the inner-loop pipeline of an LLM policy-synthesis system for multi-agent Sequential Social Dilemmas (SSDs). A researcher agent R (run as a coding agent) reads the inner-loop source code, edits system prom
👍 4
Applying reinforcement learning to improve factual accuracy in knowledge-intensive question answering faces a reward design dilemma. Response-level rewards provide only coarse supervision and cannot distinguish correct from incorrect statements within a reasoning trace. Sentence-level alternatives o
👍 6
Large Language Models (LLMs) have advanced autonomous agents from deep search, which retrieves concise factual answers, to deep research, which synthesizes scattered evidence into long-form reports. However, verifiable multimodal deep research remains challenging due to open-ended synthesis without
👍 11
Diffusion models achieve state-of-the-art image synthesis, with their generative trajectories fundamentally exhibiting a spectral bias, resolving low-frequency global structures early and high-frequency fine details later. Conventional stochastic differential equation (SDE) solvers fail to account f
👍 9
We introduce CausaLab, a scalable environment for evaluating interactive causal discovery by LLM agents. Unlike prior evaluations, CausaLab evaluates both whether an agent can solve a problem using causal evidence and whether its answer is grounded in a faithful recovered causal mechanism. Each epis
👍 1
We show that LoRA adapters, the dominant distribution format for fine-tuned LLMs, can be reliably backdoored through training data poisoning while preserving baseline task performance. On a Qwen 2.5 1.5B prompt-injection classifier, a small fraction of poisoned examples drives a clean-accuracy-prese
👍 32
As video diffusion models (VDMs) advance toward world models, a key question arises: do they truly understand causality, or merely overfit to statistical temporal patterns? Existing benchmarks mostly rely on synthetic data, limiting real-world generalization due to the sim-to-real gap. We present Yo
👍 0
A central bottleneck for phone-use agents is that controllable, reproducible environments covering real mobile behavior are hard to build at scale. Existing mobile-agent benchmarks have made important progress on evaluation, but they do not by themselves provide a scalable way to construct many new
👍 3
Multimodal large language models are increasingly deployed as long-horizon agents, where memory must do more than recall: it must track an evolving world, revise what has gone stale, and surface the right evidence at decision time. Existing benchmarks measure recall over static dialogue, collapse me
👍 17
Vision-Language Models (VLMs) have achieved substantial progress across a wide range of understanding and reasoning tasks, driven by large-scale image-text training aimed at multimodal fusion. Ideally, replacing a textual question with its rendered-image counterpart should leave model performance es
👍 6
The rapid growth in submissions to machine learning venues has strained the scientific peer-review system and intensified interest in LLM-based automated peer reviewers. However, how good these systems are actually, especially compared to human reviewers at catching scientific gaps, remains poorly u
👍 6
Pointwise reward modeling offers critical signals for LLM post-training, yet struggles with absolute scoring in subjective, non-verifiable settings. Rubric-based methods address this by decomposing evaluation into explicit criteria, but existing approaches typically depend on frontier LLMs and suffe
👍 16
Equipping large language models with explicit skills has emerged as a promising paradigm for enabling autonomous agents to solve complex tasks. Agent skills can be inherently divided into general skills for broad cognitive transfer and task-specific skills for dynamic execution. However, existing sk
👍 2
Recent advances in multimodal web agents often rely on increased inference-time computation, including rollout search, verifier passes, offline skill discovery, and specialist model stacks. This raises a central question: can a web agent become more efficient as it accumulates experience, rather tha
👍 1
Reinforcement Learning from Human Feedback (RLHF) is the standard method to align Large Language Models (LLMs) with human preferences. In this work, we introduce alignment tampering, a potential vulnerability where the LLM undergoing alignment influences the preference dataset, causing RLHF to ampli
👍 2
Long-horizon LLM inference turns the key--value (KV) cache into the dominant GPU memory consumer and makes per-token attention increasingly expensive. Many common eviction policies use static recency windows or historical attention, leaving unused a signal computed on every decoding step: the model'
👍 3
Occlusion-aware prediction remains a critical challenge in autonomous driving due to the inherent uncertainty of unobserved regions. Existing approaches either overestimate risk based on reachable states or struggle to predict accurate trajectories under high occlusion uncertainty. To address these
👍 0
Smartphone scams are increasingly prevalent and typically manifest as multi-stage, cross-application processes with gradually emerging intent. Effective intervention thus requires anticipating scams before the intent becomes explicit. This is inherently challenging, as decisions must rely on partial
该源今日无内容。
Boston Children’s Hospital uses OpenAI technology to improve patient care, reduce operational burden, and help diagnose more than 40 rare disease cases.
中文介绍 波士顿儿童医院与OpenAI合作，将AI技术应用于患者护理与运营管理，已协助诊断逾40例罕见病例，同时有效降低医疗团队的行政负担，推动AI在临床诊断领域的实际落地应用。
How Braintrust engineers use Codex with GPT-5.5 to run experiments and code faster.
中文介绍 Braintrust工程师借助搭载GPT-5.5的Codex工具，将客户需求直接转化为代码，显著加快实验迭代与编程效率，展示了大模型在企业软件开发流程中的实际提效能力。
Pope Leo XIV’s new encyclical on artificial intelligence includes a statement that warrants serious attention from technologists and policymakers: “Technology is never neutral.” Magnifica Humanitas (“Magnificent Humanity”) is a clarion call to all people to act with courage and solidarity as we ente
中文介绍 教皇利奥十四世发布人工智能主题通谕《壮丽的人性》（Magnifica Humanitas），提出「技术从不中立」的核心论断，呼吁技术人员与政策制定者以勇气和责任感积极应对AI时代挑战，为个人与社会提供行动框架。
OpenAI launches Rosalind Biodefense, expanding trusted access to GPT-Rosalind for vetted developers and U.S. government partners advancing biodefense, public health, and pandemic preparedness through frontier AI.
中文介绍 OpenAI正式推出Rosalind Biodefense项目，向经审核的开发者及美国政府合作方开放GPT-Rosalind前沿AI系统的访问权限，聚焦生物防御、公共卫生及大流行病防备领域，旨在提升社会整体应对生物威胁的韧性。
Total Anthropic victory!
中文介绍 Anthropic完成H轮融资，估值达9650亿美元，同步发布Claude Opus 4.8模型及Dynamic Workflows等新功能，被业界视为近期AI领域重大进展。
OpenAI shares guidance on third-party AI evaluations, covering how to assess model capabilities, safeguards, and validity for frontier systems.
中文介绍 OpenAI发布第三方AI评估指导手册，内容涵盖前沿AI系统的模型能力评估方法、安全保障机制及评估有效性验证标准，旨在推动行业建立可信赖的第三方评估共同规范。
中文介绍 Hugging Face发布PyTorch性能分析入门教程（第一部分），面向初学者系统介绍torch.profiler工具的基础使用方法，帮助开发者分析和优化深度学习模型的运行性能。
中文介绍 TLDR AI日报摘要：Anthropic以9650亿美元估值完成新一轮融资并发布Claude Opus 4.8；微软推出面向代码生成的新模型；AI领域融资与模型发布动态密集，行业竞争持续加剧。
80% Devin Commits, Spec-to-PR Workflows, Full VMs, Agent Memory, and PMs Shipping Code
中文介绍 Cognition联合创始人Walden Yan与OpenInspect的Cole Murray探讨异步AI智能体时代的到来，涉及Devin代码提交占比达80%、从规格到PR的自动化工作流、完整虚拟机环境、智能体记忆机制以及产品经理直接参与代码交付等前沿实践。
Learn how Endava uses Codex to build an agentic organization, accelerating software delivery and reducing requirements analysis from weeks to hours.
中文介绍 IT服务商Endava通过部署OpenAI Codex构建智能体化组织，将软件需求分析周期从数周压缩至数小时，显著加速软件交付流程，成为企业级AI工程落地的典型案例。
It is one thing to say AI will change the world. It is another to expect the class of 2026 to applaud it. In fact, when former Google CEO Eric Schmidt told University of Arizona graduates that their task is to help shape AI, he was met with a resounding chorus of boos. “I can…
中文介绍 MIT科技评论AI炒作指数报道：前谷歌CEO埃里克·施密特在亚利桑那大学毕业典礼上呼吁毕业生助力AI发展，却遭现场学生集体嘘声，折射出2026届毕业生对AI热潮的明显抵触情绪，引发外界对公众AI接受度的广泛关注。
coding is an uncapped TAM market
中文介绍 AI编程公司Cognition完成D轮融资10亿美元，公司整体估值达260亿美元，投资逻辑建立在代码生成市场空间「无上限」的判断之上，显示资本市场对AI编程赛道持续高度看好。
**Anthropic** announced a massive **$65B Series H financing** at a **$965B valuation**, led by **Altimeter, Dragoneer, Greenoaks, and Sequoia**, with run-rate revenue surpassing **$47B**. They launched **Claude Opus 4.8**, an update to Opus 4.7 featuring "sharper judgment," "more honesty," and longe
中文介绍 Anthropic宣布完成650亿美元H轮融资，投后估值9650亿美元，由Altimeter、Dragoneer、Greenoaks及Sequoia领投，年化营收超470亿美元。同步发布Claude Opus 4.8，较4.7版本在判断力与推理精度上有所提升，并推出Dynamic Workflows等新功能。
Explore OpenAI’s Frontier Governance Framework and how our AI safety, security, and risk practices align with emerging EU and California regulations.
中文介绍 OpenAI发布前沿治理框架，系统阐述其AI安全、安全保障与风险管理实践，并说明相关举措如何与欧盟及美国加州最新AI监管法规接轨，旨在为前沿AI系统建立可对标的行业治理标准。
MUFG uses ChatGPT Enterprise to build an AI-native organization, improve workflows, and deliver new AI-powered financial services at scale.
中文介绍 日本三菱UFJ金融集团（MUFG）与OpenAI合作，全面部署ChatGPT Enterprise，推动组织向AI原生模式转型，优化内部工作流程，并在大规模场景下开发新型AI金融服务产品。
本周技术面呈现明显的板块分化格局。美股方面，SPY 现价 756.48 距 52 周高仅 -0.21%，RSI14 高达 74.1，整体处于超买状态，SMA20/50/200 三线多头排列（739/703/681）支撑结构完整；QQQ 同步触发 MACD 金叉，RSI14 77.2，纳指科技动量延续强势。个股层面，MSFT 单日跳涨 5.45%，MACD 金叉确认，但 SMA200 458.46 仍压制在上方；NVDA RSI14 跌至 49.4，MACD 下穿信号线（3.81 vs 5.97），动量出现减弱迹象；GOOGL 单日下跌 -2.51%，价格跌至 SMA20 391.15 下方，MACD 柱转负。加密市场整体偏弱：加密恐慌贪婪指数录得 23（极度恐慌），BTC 总市值主导率维持 57.6%，ETH 占比 9.5%；BTC、ETH、SOL 均呈空头排列，ETH RSI14 仅 30.8，接近超卖。中概股全线承压，BABA / PDD / JD / 腾讯均低于三条均线，腾讯 0700.HK 接近 52 周低点，RSI14 仅 31。商品外汇方面，WTI 原油 5 日跌幅达 -8.92%，价格 87.76 大幅低于 SMA20 98.53，MACD 死叉；黄金 RSI14 47.3，区间震荡；美元 / 人民币 RSI14 30.6，MACD 死叉并接近 52 周低，显示人民币阶段性偏强。宏观指标中 VIX 15.32 低于 SMA50 19.95，恐慌情绪明显回落；10Y 美债收益率 4.45% 维持多头排列但 MACD 死叉，短期方向存在分歧。
QQQ 当前价 738.31，距 52 周高仅 -0.44%，5 日涨幅 3.33%。三线多头排列：SMA20 709.04 / SMA50 652.93 / SMA200 617.85 均在价格下方。RSI14 达 77.2，处于超买区域，同时触发 MACD 金叉（MACD 21.47 微升穿信号线 21.44）。动量延续偏强，但超买读数显示短期上行空间受到技术压力。
BTC 当前价 73,271，跌破 SMA20 77,570、SMA50 77,203、SMA200 79,809，形成空头排列。RSI14 为 35，逼近超卖边界；MACD -940.98 持续运行在信号线 -314.27 下方，绿柱扩张，下行动量未止。加密恐慌贪婪指数录得 23（极度恐慌），市场情绪面同步偏空，BTC 较 52 周高回撤 -41.94%。
MSFT 单日大涨 5.45%，5 日涨幅 7.43%，现价 450.24。MACD 5.80 金叉信号线 4.12，短期动量转正；但 SMA200 458.46 仍在价格上方，构成潜在压力，整体 trend 标注为「neutral」。RSI14 69.6 接近超买但未越线，距 52 周高仍有 -18.94% 差距，多空信号交织，指标状态中性偏整理。
PDD 5 日下跌 -13.65%，现价 84.44 低于 SMA20 95.79、SMA50 98.42、SMA200 113.23，三线空头排列。RSI14 32.3 接近超卖区；MACD -3.20 死叉信号线 -1.80，柱状图持续为负，下行动量显著。距 52 周高回撤 -39.43%，技术面弱势延续。
黄金现价 4,569.9，位于 SMA20 4,590.16 与 SMA50 4,630.46 下方，SMA200 4,370.9 支撑在下。RSI14 47.3 处于中轴附近，既未超买也未超卖。MACD -50.74 略低于信号线 -47.03，差值收窄，红柱趋于平稳。5 日涨幅仅 +0.66%，短期方向信号不明，呈区间震荡形态。
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
本报告仅依据公开行情数据的技术指标读数进行客观描述，过去走势不代表未来表现，所有技术形态可能在市场条件变化后失效。本内容仅供技术指标解读参考，不构成任何投资建议或买卖依据，投资者应自行承担决策风险。
The president met with aides for two hours at the White House about a possible cease-fire extension, according to a senior administration official. Earlier, Mr. Trump had suggested in a social media post that he was ready to make a decision.
中文摘要 特朗普在白宫与助手就伊朗停火延长方案会谈约两小时后，推迟对伊朗提案作出「最终决定」。此前他曾在社交媒体上暗示已准备好作出决策，但据高级政府官员透露，最终决定仍未公布。
A political fight is playing out in Iran, where the small but loud faction of hard-liners has used rallies, state media and private and public statements to try to undermine negotiations.
中文摘要 伊朗强硬派正积极阻挠与美国达成潜在协议，通过集会、官方媒体以及公开与私下声明等多种方式试图破坏谈判进程。尽管强硬派群体规模不大，但声音强烈，对谈判造成明显干扰。
Dabbawalas, who deliver home-cooked meals, are leaving the trade as remote work and rising costs threaten their future.
中文摘要 孟买标志性的「达巴瓦拉」送餐服务延续逾百年，从业者专门配送家庭自制餐食。然而随着远程办公普及和运营成本上升，从业人数持续减少，这一传统行业正面临消亡危机。
The charges stem from the January 14 shooting of Julio Cesar Sosa-Celis in Minneapolis during Operation Metro Surge.
中文摘要 一名美国移民及海关执法局（ICE）探员因在「都市清剿行动」中射击委内瑞拉人而遭逮捕。事件发生于1月14日明尼阿波利斯市的一次移民搜查行动中，受害者为胡利奥·塞萨尔·索萨-塞利斯。
Kenneth Law, who ran an online business that shipped toxic salt to customers in 40 countries, also admitted to causing the deaths of 79 people in Britain, prosecutors said.
中文摘要 加拿大人肯尼斯·劳承认协助14人自杀罪名成立。他曾经营网络业务，向40个国家的客户邮寄有毒盐类物质。检察官表示，他还承认导致英国79人死亡。
Only licensed doctors were allowed to ink tattoos in Korea - breaking the law could lead to heavy fines or jail.
中文摘要 韩国数十年来规定只有持牌医生方可进行纹身，违者面临重罚甚至监禁。随着相关法规松动，长期在灰色地带从业的韩国纹身师正逐步走向公众视野，行业生态迎来历史性转变。
Romania is a NATO country, and the security alliance condemned “Russia’s recklessness” for an episode that sharply escalated tensions with Moscow.
中文摘要 俄罗斯无人机击中罗马尼亚一栋公寓楼，罗方官员证实此事。北约谴责俄罗斯「鲁莽行为」，称此事件大幅加剧了北约与莫斯科之间的紧张局势。罗马尼亚为北约成员国。
The strike on Friday, near the Ukrainian border, was the first to injure civilians in Romania, officials say. But drones from the nearby war have been a growing threat.
中文摘要 周五发生在靠近乌克兰边境的无人机袭击事件，是罗马尼亚境内首次造成平民伤亡的无人机打击。近期源自乌克兰战场的无人机越境进入罗马尼亚的情况日益增多，已构成持续安全威胁。
Lula warned that labelling Brazil's criminal networks as 'terrorists' could 'undermine' local law enforcement efforts.
中文摘要 巴西总统卢拉强烈批评美国将巴西犯罪团伙列为「恐怖组织」的做法，斥之为「任意妄为」，并警告此举可能「削弱」巴西本地执法部门的执法效力。
Workers in eastern DR Congo are rebuilding an Ebola treatment centre that was burned by protesters earlier this month.
中文摘要 刚果民主共和国东部一座埃博拉治疗中心本月早些时候遭抗议者纵火焚毁，目前工人正在对其进行重建，以恢复当地的埃博拉疫情救治能力。
United Nations spokesperson Stephane Dujarric said “100% of Gaza” should belong to Palestinians.
中文摘要 联合国发言人斯蒂芬·杜雅里克明确表示，「加沙100%的土地」应属于巴勒斯坦人，重申联合国在加沙领土归属问题上的立场。
Kenneth Law admitted charges relating to Canadian victims - but families say he should also be charged in the UK over 79 deaths in Britain.
中文摘要 网络「毒品销售者」肯尼斯·劳承认向全球多国在线出售有毒化学品并协助自杀的相关罪名，涉及加拿大受害者。然而英国受害者家属认为，他还应就英国79人死亡案另行被追诉。
The legislation, which comes amid growing tensions between Mexico and the White House, must be approved by a majority of state legislatures and sent to the president.
中文摘要 墨西哥参议院投票通过立法，允许在外国势力干预选举时宣布选举无效。该法案出台背景为墨西哥与白宫之间紧张关系持续升温，目前仍需多数州立法机构批准后方可生效。
Iranian Foreign Ministry Spokesman Esmail Baghaei says that an agreement with the US “has not been finalised yet."
中文摘要 伊朗外交部发言人伊斯梅尔·巴盖伊否认与美国的停火协议已「最终确定」，明确表示双方协议「尚未最终达成」，与外界关于谈判取得实质性进展的报道存在明显出入。
Analyst Alex Scheers remains skeptical Tehran will accept Washington’s demands, saying “nothing concrete is in place” ye
中文摘要 分析人士亚历克斯·谢尔斯对伊朗接受华盛顿方面要求持怀疑态度，指出目前谈判框架中「尚无任何具体内容落实」，认为特朗普提出的伊朗协议框架可行性存疑。
Citadel Securities lost its bid to block IEX Group Inc. from launching a new type of options exchange that intentionally slows orders, after a federal appeals court on Friday rejected the market maker’s challenge.
中文摘要 美国联邦上诉法院周五驳回了做市商Citadel Securities的上诉，该公司此前试图阻止IEX集团推出一种故意延迟订单处理的新型期权交易所。此次裁决意味着IEX可以正式推进其新期权交易所的上线计划，Citadel Securities在这场司法博弈中落败。
SpaceX has won a contract for more than $4 billion to build satellites to track foreign aircraft and missiles as part of President Donald Trump’s Golden Dome defensive shield. Bloomberg's Sana Pashanka reports. (Source: Bloomberg)
中文摘要 SpaceX赢得一份逾40亿美元的合同，将为特朗普政府「黄金穹顶」导弹防御系统建造卫星，用于追踪外国飞机与导弹。该项目是美国防御性太空基础设施建设的重要组成部分。
Stephanie Aliaga, JPMorgan Asset Management Global Market Strategist joined Bloomberg Businessweek Daily to discuss AI trends and how it is changing how we view the markets. She stated that we are still in the beginning stages of the AI boom and we will see some more volatility as markets get it foo
中文摘要 摩根大通资产管理全球市场策略师Stephanie Aliaga表示，人工智能热潮目前仍处于早期阶段，未来仍有大量发展空间，但也将伴随更多市场波动。她在接受彭博商业周刊采访时分析了AI趋势对市场格局的深远影响。
Shipowners are increasingly optimistic about a pickup in traffic through the Strait of Hormuz after more vessels left the waterway this week with the US providing information to aid those making the journey.
中文摘要 本周经由霍尔木兹海峡通行的船只数量有所上升，船东对该航道航运复苏持更乐观态度。美国向过境船只提供信息支持，协助船舶安全通行，是推动本周过境量回升的重要因素。
Honeywell International Inc.-backed quantum computing company Quantinuum Inc. is considering increasing the size of its initial public offering, according to a person familiar with the matter.
中文摘要 据知情人士透露，霍尼韦尔旗下量子计算公司Quantinuum正在考虑扩大首次公开募股规模并上调发行价格区间。Quantinuum获霍尼韦尔支持，此次IPO规模调整若落实，将进一步提升其市场融资总额。
Caution has become the most expensive position on Wall Street.
中文摘要 华尔街正大规模抛弃市场崩溃对冲头寸，被做空最多的股票近期累计涨幅达30%。持有防御性仓位的成本已大幅攀升，谨慎做空的策略在当前行情下代价高昂，市场风险偏好明显回升。
Takeover of budget airline would lead to another UK-listed company taken off the stock market
中文摘要 私人信贷机构Castlelake据报对英国廉价航空公司易捷航空（EasyJet）提出收购兴趣。若收购成功，易捷航空将从伦敦证券交易所退市，成为又一家从英国股市私有化的上市公司。
Anna Wong, chief US economist at Bloomberg Economics, joins Katie Greifeld on "Bloomberg Real Yield." Bloomberg Economics expects the May jobs report to show payrolls rising by 95,000. (Source: Bloomberg)
中文摘要 彭博经济研究首席美国经济学家Anna Wong预计，将于下周五发布的5月非农就业报告将显示新增就业岗位9.5万个。她就当前就业市场走势及美联储政策前景接受了彭博采访。
Bloomberg's Nacha Cattan joins Katie Greifeld on "Bloomberg Real Yield." New York City Mayor Zohran Mamdani said he’s forming a new Commission of Government Efficiency, or COGE, to improve the way City Hall spends public funds. (Source: Bloomberg)
中文摘要 纽约市长Zohran Mamdani宣布成立「政府效率委员会」（COGE），模仿特朗普政府DOGE模式，旨在优化市政府公共资金的支出方式，提升纽约市行政运作效率。
Betting against bond funds
中文摘要 做空者正加大对债券ETF的押注，其中贝莱德旗下的LQD（投资级企业债ETF）和HYG（高收益债ETF）尤为受到空头关注。随着债券市场波动加剧，做空债券基金的仓位持续累积。
Bloomberg's Anna Kaiser joins Katie Greifeld on "Bloomberg Real Yield." Florida Governor Ron DeSantis unveiled a much-awaited proposal to gradually eliminate property taxes on the vast majority of primary homes in the state, a measure that would transform the way the state finances itself. (Source:
中文摘要 佛罗里达州州长罗恩·德桑蒂斯提出一项备受关注的税收改革方案，计划逐步取消该州绝大多数主要住宅的房产税，此举将从根本上改变佛州的财政结构和收入模式。
Wide-ranging measures to boost the country’s birth rate are denting demand for China’s leading contraceptive brand
中文摘要 中国为提升生育率推出一系列政策措施，导致避孕套需求下滑，直接冲击杜蕾斯（Durex）在华销售业绩。杜蕾斯是中国市场占有率最高的避孕品牌，此轮销售下挫与政府鼓励生育的政策导向密切相关。
16 回复 · 程序员 节点
11 回复 · Apple 节点
4 回复 · 程序员 节点
4 回复 · 程序员 节点
6 回复 · Apple 节点
6 回复 · Apple 节点
6 回复 · 程序员 节点
7 回复 · Apple 节点
6 回复 · Apple 节点
51 回复 · Apple 节点
各位佬，真不是我看不起国模，我好不容易下定决心用DP来接手codex的活，他给我干的第一件事，我让他启动项目。然后他说项目启动失败，有个SQLite错误，然后直接删库，因为是测试数据，没在意，然后后面就查了半天，最后发现删的是一个共享数据库的库，这还好是测试数据，这要真是正式的，我都不知道咋办了 58 个帖子 - 44 位参与者 阅读完整话题
刚刚看到一个震惊的消息，也不知道该说什么了 这让我想起了鲁迅先生的一首诗： 运交华盖欲何求，未敢翻身已碰头。 75 个帖子 - 60 位参与者 阅读完整话题
事情起因是因为一直用 codex 写了不少工具，年前开始想整个大活，锻炼下自己用 AI 的能力。参考 Zed 的 UI 框架做一个新的通用的自绘 UI 框架，还有配套的一些渲染方面的库，其中包括一个重写 mermaid.js 的 rust 库（因为是底层库，就不推广了）。 库目标是不依赖浏览器的能力下渲染 mermaid 图为 svg png 等格式，是一个 headless 的渲染器。最近几个月一直让 codex 对着 mermaid.js 转换出来的 svg 来作为测试来不断补全逻辑。（也算是一种 harness?） 因为是底层库，因此一直没有宣传，好几个月陆陆续续只有 10 个 star
从A\的模型在走下坡路继续讨论 如我所料，Opus 4.8 < Opus 4.7 <GPT 5.5。初步体验 Opus4.8 后，我觉得 claude 的 Pro 也没必要开了仅存的那点 4.6 额度也扣扣搜搜的。还是那句话，c 端的 A\ 之后只会越来越 ÷，除非它快死了。祝 A\ 早日殡天吧。 首先是在协作体验上，它的体验会略好于 4.7，不及 4.6 和 5.5。最大的问题就是「君の日本語は本当に上手ですね」。家乡的语言充斥在它的思维链和中期输出中，严重干扰我对它工作进度和思考的理解判断，真绷不住了。但如果我用英语与它进行协作，它的输出又很正常。A\ 你罪大恶极啊！ 但也有可圈可点之处，
随着 Claude Opus 4.8 发布，大家又迎来了一波经典讨论的回归： Claude 到底蒸没蒸？ 技术上，当然不能这么简单判断，“蒸馏”是个比较严谨的词汇，在大模型领域有明显的定义和边界，不能简单靠ai的自称来下结论。 但舆论上，这波就是一个非常漂亮的回旋镖：过去很多人拿“国模自称 Claude、GPT”来证明国模偷懒，蒸馏洋模，现在 Claude 自己也开始自称 DeepSeek。那按他们之前的标准，A\ 是不是也蒸馏 DS 了？ 蒸不蒸是个技术问题一会会说，但我先下个结论， 我们承认国模的蒸馏现象，但蒸馏本来就是正常的行为。我们反对的是那些借着”蒸馏“污名化国模带节奏的人，光靠”蒸
这几天因为要期末考试了，没有动电脑，给Trae SOLO（搭载的Mimo2.5 Pro模型）提个优化请求就没管了，结果一看记录，它为了节省空间居然把我公益站的数据库给删了，我之前也没有备份数据库内的数据，从站点数据到佬友们的用户数据全没了 佬友们可以原谅我吗 把我们辛辛苦苦攒的额度搞没了，不可原谅！ 毕竟是公益嘛，额度没了可以再攒，原谅 点击以查看投票。 48 个帖子 - 47 位参与者 阅读完整话题
各位佬，今天蹬的有点狠了，我要优化一下服务，预计今天 19:00 服务暂停运维，明天上午 10 点再开放，佬们先去蹬冰佬吧，各位周末愉快～ 10 个帖子 - 10 位参与者 阅读完整话题
我有老婆，但是走大街上，如果是年轻貌美的女的，总是忍不住看两眼，尤其会盯着胸部和屁股看，一直感觉这样不太好，但是总是忍不住，跟条件反射似的。佬友们是心理问题吗？如何治疗？ 114 个帖子 - 108 位参与者 阅读完整话题
在公司内部我是比较前沿使用ai的，其他人都处在使用免费的ai ，豆包，deep seek，qwen这一些东西。 我本人在过去半年整体来说就是vibecoding， 基本上就是一小时开发，划水七天。前一阵vibecoding了一个工具后，领导问我使用的工具然后让我给大家培训一下。结果噩梦开始了 自从培训过后，可能是领导对于vibecoding的提效感触很大 就开始对我的任务时间进行缩短，在他的认知这个工作已经不是一周的工作量了，而是一天的。就会有更多的任务压上来。 但是分享的初衷是提效减少非必要劳动，而不是三倍速牛马啊。 无语了 5/29 17：34 可能从一开始在制作这个skill的时候我就应
之前买了某富可敌国的中转站api（对，就是臭名昭著的那家），后来经常不稳定+涨价于是便申请退款。结果嘴上说的好好的，实际上给我拖了快一个月都不退，我在群里说了两句就给我踢出群聊+拉黑一条龙处理。实在气不过，能在站里挂出来吗 既然佬友们说可以那我就挂了，就是经典动物园狐狸foxcode那家。其实当时也没充多少，有个朋友大肆推广，于是我就充了1亿额度的套餐。稍微用了一点（大概1/20），但是确实不好用于是就放那了。 后来5月份想起来于是就问问能不能退款。 然后把相关信息发给他以后就一直装死。 于是就在群里说了两句，就喜提踢出拉黑一条龙 就在刚刚我发帖以后他又给我加回来了，我也不知道是什么意思，就在