TypeScript · ★ 8,798 · 🍴 815 · 📈 2,483 stars today
A self-improving RLM agent for coding workflows and long-running autonomous tasks.
中文介绍 prime-agent 是一个基于强化学习（RLM）的自主代理，专为编码工作流和长期任务设计。它能在执行代码生成、重构、测试等开发任务时不断自我改进，积累经验并优化策略。适合需要无人值守完成复杂开发流程的工程师或团队，尤其是自动化多步骤工程场景。
JavaScript · ★ 84,516 · 🍴 9,080 · 📈 778 stars today
Production-grade engineering skills for AI coding agents.
中文介绍 Agent Skills 是 Addy Osmani 整理的生产级工程技能包，为 AI 编码代理提供可复用的实践指导。它解决了通用模型缺乏真实项目规范的问题，包含代码评审、测试、调试等技能，可直接加载到 Copilot、Cline 等代理中使用，帮助软件团队提升代码质量和开发效率。
Roff · ★ 77,899 · 🍴 17,632 · 📈 591 stars today
所有小初高、大学PDF教材。
中文介绍 ChinaTextbook 收录了从小学、初中、高中到大学的全学科 PDF 教材，覆盖面广且完全免费。它解决了学生和自学者获取正版电子教材困难的问题，以静态 PDF 形式提供，支持离线下载与阅读。适合备考、自修、教学等场景，是开源中文教材的重要集合。
Python · ★ 16,704 · 🍴 1,355 · 📈 481 stars today
Agent Skills for Google products and technologies
中文介绍 google/skills 是谷歌官方发布的 Agent Skills 合集，帮助 AI 代理调用 Google 产品和技术，如搜索、地图、Workspace 等。这些技能以结构化的指令和工具定义嵌入代理系统，使代理能完成真实世界任务，适合构建智能助理、自动化工作流和多云应用的开发者使用。
Shell · ★ 209,955 · 🍴 18,138 · 📈 1,354 stars today
Skills for Real Engineers. Straight from my .agents directory.
中文介绍 mattpocock/skills 是 TypeScript 专家 Matt Pocock 从个人 .agents 目录中整理的实用技能集，专注于真实工程需求。它提供代码审查、调试、架构建议等可复用技能文件，让 AI 编码代理加载后获得更专业的工程经验，适合前端、全栈项目中希望提升代理实战能力的开发者。
Python · ★ 23,954 · 🍴 1,836 · 📈 467 stars today
The authentication glue you need.
中文介绍 authentik 是一个开源身份认证与访问管理（IAM）平台，提供单点登录、多因素认证、授权策略等核心能力，并能与 LDAP、OAuth2、SAML 等协议集成。它解决企业内部应用认证分散问题，适合需要统一身份认证的组织，尤其兼容 Kubernetes 和云原生架构，可简化安全治理。
Python · ★ 96,446 · 🍴 18,650 · 📈 126 stars today
TradingAgents: Multi-Agents LLM Financial Trading Framework
中文介绍 TradingAgents 是一个基于多代理 LLM 的金融交易框架，模拟多个 AI 代理协作、辩论与三阶段推理来做出投资决策。它利用大语言模型分析新闻、市场数据和情绪，以对抗机制减少偏见，适合量化研究者、算法交易员和金融 NLP 开发者进行策略研究与回测。
Java · ★ 51,843 · 🍴 11,180 · 📈 93 stars today
Google core libraries for Java
中文介绍 Guava 是 Google 官方的 Java 核心库，提供集合、缓存、并发、字符串处理、哈希、事件总线等常用工具和 API，减少样板代码并简化开发。它解决了 JDK 生态中基础设施不足的问题，被全球大量生产级系统采用，适合任何使用 Java 构建中大型应用的后端团队。
C++ · ★ 64,971 · 🍴 3,104 · 📈 79 stars today
Truly independent web browser
中文介绍 Ladybird 是一款真正独立的 Web 浏览器，基于自研布局引擎和 JavaScript 引擎，不依赖 Chromium、Firefox 或 WebKit。它旨在保持 Web 的开放性与多样性，提供高性能渲染，并支持多个桌面平台。适合浏览器开发者、开源爱好者和关注技术独立性的用户参与开发与测试。
Rust · ★ 2,539 · 🍴 71 · 📈 432 stars today
self-hosted, distributed Durable Objects
中文介绍 celld 是 Deno 团队推出的自托管分布式 Durable Objects 实现，让开发者在自有基础设施上运行持久化、有状态的计算单元。它解决无服务器应用中状态跨节点一致性问题，提供类似 Durable Objects 的 API，适合构建实时协作应用、协同编辑器和边缘计算场景。
★ 693 · 🍴 880 · 📈 59 stars today
DevOps Interview Guide
中文介绍 DevOps 面试指南是一个面向求职准备的考点汇总仓库，覆盖 CI/CD、容器、Kubernetes、监控和 Linux 等核心技能，并附有常见问题与答案。它帮助候选人系统复习 DevOps 知识，也为团队招聘提供考核参考，适合准备运维、SRE 或 DevOps 岗位的技术人员。
Kotlin · ★ 49,860 · 🍴 8,252 · 📈 161 stars today
翻墙-科学上网
中文介绍 fanqiang 是一套突破网络访问限制、实现自由上网的资源集合，汇总了代理工具、配置教程、服务器推荐和翻墙原理说明。它面向有合规需求或研究目的的用户，提供可执行文件与脚本，覆盖 Windows、macOS、Linux 等平台，常用于访问开放互联网信息。
👍 218
High-quality long-horizon training data for terminal agents is expensive to produce, often costing hundreds to thousands of dollars per task, because each task must keep the instruction, environment, reference solution, and verifier mutually consistent. Human authoring does not scale, and direct gen
中文介绍 针对终端智能体长时程任务训练数据昂贵且难以保证指令、环境、参考解与验证器一致的问题，提出递归合成方法，自动生成高质量长时程任务数据，缓解人工编写不可扩展的瓶颈，降低任务级数据成本。
👍 75
Reinforcement learning (RL) with verifiable rewards constructs trajectory-level advantage estimates, yet it often fails to credit the few pivotal decisions that determine outcomes in long-horizon, multi-turn agentic tasks. Recent work introduces privileged self-distillation for credit assignment, pr
中文介绍 针对长时程多轮智能体任务中轨迹级优势估计难以对少数关键决策进行信用分配的问题，提出AgentOPSD递归自蒸馏方法，通过递归蒸馏关键决策信号改进智能体强化学习，提升信用分配的准确性。
👍 63
Long-horizon search agents must make multiple sequential actions (steps) to search, retrieve, verify, and integrate evidence to reach a final answer. However, existing methods for training these agents typically treat all steps within a trajectory uniformly during both supervised fine-tuning (SFT) a
中文介绍 长时间搜索智能体需通过搜索、检索、验证和整合证据等多步行动得到答案，现有SFT与RL常对所有步骤一视同仁。提出ABSeeker，利用答案回溯进行信用分配，区分关键步骤，提升长时程搜索智能体训练效果。
👍 60
Computer-using agents (CUAs) are advancing rapidly across the digital world. A CUA trajectory records the agent's actions, states, and reasoning. Verifying whether it fulfilled the task instruction is central to CUA evaluation, data curation, and reinforcement learning. Neither human-written verifie
中文介绍 提出OSReward，为跨平台计算机使用智能体（CUA）的奖励模型建立标准化评测。聚焦CUA轨迹是否完成任务指令的验证问题，覆盖动作、状态与推理，支撑CUA评估、数据筛选和强化学习。
👍 58
Short segments of perceived speech can be retrieved from non-invasive magnetoencephalographic (MEG) recordings by deep networks trained with a CLIP-style objective against wav2vec 2.0 audio embeddings. Yet their weights do not map onto electrophysiological quantities, and it remains unclear which sp
中文介绍 针对深度网络从MEG信号解码语音但权重无法对应电生理量、驱动特征不明确的问题，采用可解释性分析定位皮层来源和关键刺激特征，揭示CLIP式训练结合wav2vec 2.0嵌入所依赖的语音表征。
👍 50
Generating large-scale, freely explorable 3D worlds from open-ended text remains challenging because a system must jointly maintain global spatial coherence, rich local content, and explicit assets suitable for downstream editing and reuse. We present WorldClaw, a fully agentic, coarse-to-fine frame
中文介绍 提出WorldClaw，一个完全智能体化的从粗到细框架，从开放式文本生成大规模可自由探索的3D世界，兼顾全局空间连贯性、丰富局部内容以及可供编辑复用的显式资产。
👍 37
Spatial intelligence is fundamental to embodied agents, yet existing benchmarks focus on local spatial perception from single or few viewpoints, overlooking global spatial awareness over continuous, long-horizon visual streams. To address this limitation, we introduce the Global-Spatial-Temporal Ben
中文介绍 针对现有空间基准仅评测单或少数视角的局部感知、忽视连续长时程视觉流中全局空间意识的问题，提出GST-Bench，系统评估视频大模型能否发展出全局空间-时间认知能力。
👍 34
Training large language model agents for long-horizon tool use typically relies on interactions with real or synthesized executable environments, whose construction and verification are costly, or on external simulators that are difficult to ground. We introduce EnvACE, an agentic reinforcement lear
中文介绍 提出EnvACE，通过世界排演（world rehearsal）内化环境动态，用于长时程工具使用智能体的强化学习，减少对真实或合成可执行环境及外部模拟器的依赖，降低构建与验证成本。
👍 33
Multimodal large language models excel at passive perception but struggle with complex visual cognitive tasks requiring multi-step temporal reasoning. This degradation largely stems from the inherent ambiguity of language-based reasoning, which often fails to accurately articulate continuous visual
中文介绍 多模态大模型在需要多步时间推理的视觉认知任务上表现不佳，源于语言化推理难以准确表达连续视觉过程。提出ChronoVision，通过潜在状态重建进行时间推理，减少语言歧义，提升复杂视觉任务表现。
👍 31
Unified multimodal retrieval aims to identify candidates that satisfy complex user intent expressed through heterogeneous inputs. Although Large Vision-Language Model (LVLM)-based retrievers are efficient and scalable, directly encoding raw multimodal inputs often misses fine-grained discriminative
中文介绍 针对统一多模态检索中直接编码原始输入难以捕捉细粒度判别信息的问题，提出基于硬负样本的检索中心思维链（Retrieval-Centric CoT），让LVLM检索器从失败中学习，提升复杂用户意图下的检索能力。
👍 28
Economic World Models (EWMs) are generative economic models that simulate how economies evolve from within by modeling heterogeneous agents, their beliefs and actions, and the market and institutional mechanisms through which their interactions produce aggregate outcomes. This paper develops an impl
中文介绍 提出经济世界模型（EWM）的系统蓝图，从经济智能体迈向智能体经济，通过建模异质智能体的信念、行动以及市场与制度机制，生成式模拟经济体从内部演化的过程。
👍 28
As LLMs are increasingly deployed within agentic systems, their capabilities depend not only on the model weights but also on the harness: the prompts, tools, control flow, memory, and orchestration code surrounding them. This makes automated harness optimization -- the iterative and evaluation-guid
中文介绍 提出HarnessOpt-Bench，评估LLM在智能体系统中对harness（提示、工具、控制流、内存、编排代码）进行自动化优化的能力，指出模型能力不仅取决于权重，还取决于外围编排的迭代优化。
👍 25
Data agents enable natural-language analytics over organizational workspaces, where relevant evidence may be scattered across databases, structured files, long documents, and multimedia. Existing benchmarks largely isolate structured querying, retrieval, or open-ended analysis, leaving heterogeneous
中文介绍 提出DataSpace基准，面向数据库、结构化文件、长文档和多媒体等异构工作空间，评测数据智能体基于自然语言进行可验证分析的能力，弥补现有基准在结构化查询、检索与开放分析上的割裂。
👍 24
Modern Greek is absent from NVIDIA's Nemotron retrieval models and from major multilingual retrieval benchmarks, despite being important for retrieval-augmented generation (RAG) in legal, energy, financial, and medical applications. We present an end-to-end adaptation of the Nemotron retrieval stack
中文介绍 针对现代希腊语在NVIDIA Nemotron检索模型和多语言基准中缺失的问题，端到端适配Nemotron检索栈，包括语料挖掘、检索适配与生成接地，覆盖法律、能源、金融、医疗等专业领域的RAG应用。
👍 24
On-Policy Distillation (OPD) is emerging as a promising alternative to reinforcement learning for LLM post-training, yet its effectiveness in multilingual settings remains underexplored. We study OPD and its advanced variant, On-Policy Delta Distillation (OPD^2), for mathematical reasoning in Englis
中文介绍 研究On-Policy Distillation（OPD）及其高级变体On-Policy Delta Distillation（OPD²）在多语言数学推理中的效果，作为RL后训练的替代方案，检验其在英语等多语言设定下的有效性。
👍 24
Agent self-evolution updates an agent's persistent state from prior experience and reuses it to solve related tasks more effectively. Evaluating self-evolution is difficult: existing benchmarks provide limited coverage of economically valuable task domains, do not always design training and test tas
中文介绍 提出GDPevo，面向真实商业任务评估智能体自我进化能力，即从先前经验更新持久状态并复用于相关任务，弥补现有基准在经济价值任务覆盖和训练测试设计上的不足。
👍 18
Computer-use agents pay full frontier inference to re-derive routines their user has already performed, because an agent's memory today records what the user said, not what the user did. We compile passively captured screen activity into agent memory with a deterministic, zero-model pipeline: it seg
中文介绍 针对电脑使用智能体重复推导用户已执行操作而浪费推理的问题，提出Activity Frames，用确定性零模型流水线将被动捕获的屏幕活动编译进智能体记忆，记录用户行为而非言语，支持记忆与回放。
👍 18
Understanding 3D scenes is fundamental to embodied intelligence, requiring joint reasoning over heterogeneous information from multiple modalities, including visual and geometric cues. However, the relevance of these modalities often varies across queries. Existing Multimodal Large Language Models (
中文介绍 针对3D场景理解中视觉与几何等模态相关性随查询而变化的问题，提出SmartMage，动态编排多模态信息，使多模态大模型按查询选择最相关模态，从而提升3D场景理解能力。
👍 18
This technical report presents K-EXAONE 2.0, an open-weight multilingual foundation model developed by LG AI Research as a step in our effort toward global frontier-scale foundation models. Rather than training from scratch, we upcycle K-EXAONE and expand its architecture, yielding a Mixture-of-Expe
中文介绍 介绍LG AI Research开发的开源多语言基础模型K-EXAONE 2.0。采用upcycling策略在已有K-EXAONE基础上扩展架构，得到混合专家（MoE）模型，而非从零训练，向全球前沿规模模型迈进。
👍 16
Video object removal must eliminate not only the target object but also its induced effects while maintaining high-fidelity and spatiotemporally coherent restoration. Existing methods mainly learn object-effect correspondences implicitly from predefined effect categories and fixed data distributions
中文介绍 提出EffectLearner，通过世界感知的对象效果推理实现真实世界视频对象移除，在删除目标物体的同时消除其诱发效果，避免依赖预定义效果类别和固定数据分布，保持高保真时空一致修复。
@tetsuoai · 236.6K 粉丝 · 493.9K 阅 · 851 赞 · 138 转
Grok Imagine crossed a line this week. Imagine Video 1.5 shipped last month with better motion, physics, and audio. On July 31 xAI added the piece that makes actual filmmaking possible: image and
中文介绍 Grok Imagine 本周实现突破：上月发布的 Imagine Video 1.5 改进了运动、物理和音频，7月31日 xAI 又加入图像功能，使完整电影制作成为可能。内容涵盖从构思到成片的完整流程。
@milesdeutscher · 675.2K 粉丝 · 103.7K 阅 · 514 赞 · 29 转
I've officially cloned myself with Claude. This isn't another generic AI second-brain article; what I'm about to reveal is the exact system I've used to automate 80%+ of my life using AI. Every
中文介绍 作者用 Claude 搭出一套可复制的自动化系统，声称已克隆自己、自动化 80% 以上的生活。他强调这不是又一篇 AI 第二大脑文章，而是精确到配置与流程的完整方案，适合想系统性提效的人。
@hwchase17 · 124.4K 粉丝 · 86.0K 阅 · 506 赞 · 68 转
Today we’re launching Managed Deep Agents - the easiest way to build, run and deploy production agents. We have a full release blog and docs if you want to read more. In this article I want to talk
中文介绍 LangChain 发布 Managed Deep Agents，称其为构建、运行和部署生产级 agent 最简单的方式。文章重点论证：托管式 agent 是 agent 构建领域下一个大趋势，并附完整发布博客与官方文档。
@vasuman · 88.5K 粉丝 · 80.7K 阅 · 569 赞 · 33 转
You’re already in the top 1% of AI users. Yes, there’s a gap between you and the folks on the frontier. The crazy kids running 20 terminals simultaneously with a knowledge base that rewrites itself
中文介绍 作者指出 AI 采用是个神话：你已经属于前 1% 的 AI 用户，但与前沿玩家仍有代差——那些同时开 20 个终端、拥有自动重写知识库的人。真正极致的 AI 工作流远未普及。
@jino_rohit · 6.9K 粉丝 · 20.3K 阅 · 519 赞 · 59 转
This blog is my attempt to learn how to profile an llm served using sglang and its in built torch profiler integration that comes out of box. The idea here would be try and understand when I hit
中文介绍 这篇博客是作者的学习记录：用 SGLang 内置的 Torch Profiler 对 LLM 推理服务进行性能剖析，逐步定位瓶颈，并对 SGLang 自带的 profiler 集成做了实操演示，是推理调优的一份实用入门参考。
a quiet day lets us find some connections among recent themes
OpenAI is sharing preliminary cybersecurity evaluations for Astra and the steps we’re taking to strengthen safeguards and security controls.
Discover how HSP GRUPPE uses ChatGPT Enterprise to boost productivity, improve work quality, and create more capacity for tax advisory and client service.
**OpenAI** escalates its upcoming **Astra** model to "critical" cyber status due to significant advancements in agentic coding and cybersecurity, pausing some activities to strengthen controls. The "Hugging Face incident" highlights persistent multi-agent coordination failures involving externalized
The Inference Inflection is HEATING up.
ChatGPT introduces improved GPT-5.6 Sol with better accuracy and consistency, plus expanded access for free users and unlimited everyday chats with GPT-5.6 Luna.
OpenAI and the American Psychological Association advance evidence-based guidance, resources, and safeguards for responsible AI use and youth mental health.
**Meta's Muse Spark 1.2** rapidly rose to frontier-tier with top 5 ranking on Vals Index at **$0.69/test**, being **3x cheaper than Kimi** and **10x+ cheaper than Fable, Opus, and 5.6 Sol**. It achieved **gold-medal-level performance in five STEM Olympiads** with perfect theory scores in APhO and IP
The end of an era.
New OpenAI Signals data shows how people use ChatGPT worldwide, with country-level insights on adoption, usage trends, and evolving behavior.
美股技术面整体偏强：SPY、QQQ与NVDA均运行于20/50/200日均线上方，SPY距52周高点仅-0.46%，RSI 66；但MSFT RSI 78.1进入超买，TSLA与META仍处空头排列。加密市场情绪偏谨慎，恐慌贪婪指数30（恐慌），总市值2.30T，BTC主导率56.7%；BTC/ETH短期站上20日与50日均线，但仍低于200日均线，BTC RSI 54.6且MACD金叉，属中性修复。中概股普遍站上短中期均线但受制于200日均线，JD RSI 74.3超买，整体中性略偏多。商品与外汇分化：黄金单日涨3.72%、5日涨8.66%动量强劲；WTI原油5日跌7.67%且MACD死叉；美元兑人民币RSI 26.9超卖，DXY跌破20日与50日均线。VIX连续下行至14.9，避险指标回落。
追踪标普500的SPY现价773.26，距52周高点仅-0.46%，RSI 66，价格高于SMA20（750.17）、SMA50（747.19）与SMA200（702.97），呈多头排列；MACD柱6.19高于信号2.85，且触发金叉。5日涨3.51%，短线动量偏强，技术setup整体偏上行。
NVDA现价223.96，5日涨11.56%、1日涨2.27%；价格位于SMA20（206.82）、SMA50（206.07）与SMA200（193.98）上方，多头排列；RSI 64.4，MACD柱3.05高于信号0.55，MACD金叉。距离52周高点-5.32%，短线趋势动量健康，技术setup偏上行。
TSLA现价328.58，虽1日涨2.83%，但价格低于SMA20（343.33）、SMA50（380.44）与SMA200（408.80），呈空头排列；RSI 41.4位于50下方，MACD值-19.76虽略高于信号-20.78，但均在零轴下方；距离52周高点-34.13%，整体技术setup偏下行。
WTI原油现价78.18，5日跌7.67%，价格跌破SMA20（81.82）与SMA50（80.47），仅高于SMA200（76.28）；RSI 46.7，MACD-0.45低于信号0.27，MACD死叉；52周高点回撤-34.57%，短中期趋势动量偏弱，技术setup偏下行。
BABA现价128.41，1日涨1.26%、5日涨5.04%，价格高于SMA20（118.99）与SMA50（113.81），但仍低于SMA200（140.09）；RSI 67.9接近超买，MACD柱4.29高于信号2.84，方向中性略偏多；因长周期均线仍压制，技术setup判定为中性。
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
以上内容仅为基于公开行情数据的技术指标状态描述，过去走势不代表未来表现，不构成任何涨跌预测或交易建议；本报告仅供技术指标解读参考，读者应独立判断并自行承担风险。
Government promises harsh response following explosives attack on Pan-American Highway in southwest of country.
中文摘要 哥伦比亚强硬派总统就职后，该国西南部的泛美公路发生汽车炸弹袭击，引发震动。政府承诺严厉回应，但报道未提及具体伤亡情况。
The Bald Range fire has burned more than 25,000 acres in British Columbia. At one point, the fire spread more than 2,000 acres in about an hour, officials said.
中文摘要 加拿大不列颠哥伦比亚省Bald Range山火过火面积已超2.5万英亩，迫使1.8万人撤离。官员称，山火一度在一小时内蔓延逾2000英亩。
Yolŋu hip-hop artist dominates, while Butchulla duo and cousins Birdz and Fred Leone win artist of the year and Drifting Clouds takes home best new talent Baker Boy has won album, song and film clip of the year at the 2026 National Indigenous music awards (Nimas), paying tribute to his late grandmot
中文摘要 2026年澳大利亚全国原住民音乐奖（Nimas）揭晓，尤尔古族嘻哈艺人Baker Boy包揽年度专辑、歌曲和音乐录影带三项大奖；Birdz与Fred Leone获年度艺人，Drifting Clouds获最佳新人。
The Bald Range wildfire in British Columbia, still considered "out of control", has spread to more than 36 sq miles (95 sq km).
中文摘要 加拿大不列颠哥伦比亚省Bald Range野火已蔓延至逾36平方英里（95平方公里），规模扩大一倍，仍处于“失控”状态，迫使数千人逃离家园。
Some experts say the deal precipitates a larger strategic shift in Middle East while others say it reinforces US goals.
中文摘要 巴基斯坦、沙特和土耳其签署防务协议，引发对美国战略影响的讨论。有专家认为该协议预示中东更大战略转变，也有观点称其强化了美国目标。
An agreement over the Strait of Hormuz is still far away. Tim Wilkins, who works for a tanker industry group, says the ships stuck around the waterway are in limbo, and transiting is still dangerous.
中文摘要 霍尔木兹海峡协议仍遥不可及。油轮行业组织人士蒂姆·威尔金斯（Tim Wilkins）表示，滞留该水道的船只处于悬置状态，过境依然危险。
Todd Blanche, President Trump’s former lawyer, is now US attorney general.
中文摘要 美国总统特朗普的前律师托德·布兰奇（Todd Blanche）现已出任司法部长。半岛电视台报道了这一消息，但未提供更多细节。
Tehran urges US to ‘correct its behaviour’ as UAE says one of its ships targeted by Iranian missile Iran has issued tough new demands for reopening the strait of Hormuz, while the United Arab Emirates said one of its ships was targeted by an Iranian missile, as a deal to open the strategic waterway
中文摘要 伊朗就重新开放霍尔木兹海峡提出强硬新要求，敦促美国“纠正其行为”。阿联酋称其一艘船只遭伊朗导弹袭击，目前相关协议仍难达成。
Wildfire activity has increased in British Columbia, with thousands evacuated from the Okanagan Valley, weeks after smoke from blazes in Ontario choked the East Coast.
中文摘要 加拿大山火活动增加：不列颠哥伦比亚省奥肯那根谷地数千人疏散；数周前，安大略省山火产生的浓烟曾笼罩东海岸。
Hundreds of people attended a seven-day retreat in the Turkish countryside seeking transcendence through sema, the spiritual practice of whirling.
中文摘要 数百人参加土耳其乡村为期七天的静修活动，通过旋转舞（sema）这一灵性修行方式寻求超脱体验。该活动向所有人开放。
Saudi Arabia turns to Turkiye and Pakistan for defence ties, raising questions overreliance on the US.
中文摘要 沙特阿拉伯转向土耳其和巴基斯坦寻求防务合作，引发对其过度依赖美国的质疑，也使得麦加协议面临的挑战成为关注焦点。
Police in the French capital tighten rules for riders after a sharp increase in fatal accidents.
Warning of increased activity as patrols spent 21 days in July shadowing Moscow’s ships in Dover Strait and North Sea The Royal Navy has warned of increasing Russian activity around the UK’s waters after it spent 21 days in July monitoring Moscow’s ships. Naval operations to monitor Russian vessels
Advocates and legal bodies want the debt to be made ‘out of bounds’ given how hard it is to determine its accuracy Get our breaking news email, free app or daily news podcast Centrelink is pursuing $4.93bn in unpaid debts, with the oldest dating back more than 40 years, despite the government agreei
Exclusive: Workers’ rights experts raise ‘alarm bells’ over app requiring ABNs for jobs they say lack worker autonomy Get our breaking news email, free app or daily news podcast Concerns have been raised that workers using an increasingly popular Australian hospitality gig-work app may be missing ou
The news doesn’t stop when markets close. Hosts David Gura, Christina Ruffini and Lisa Mateo bring clarity, context and a bit of humor to the weekend’s biggest headlines, LIVE from New York. Joined by Oxford Centre on Migration, Policy & Society Senior Researcher Myriam Cherti, Merrill & Bank of Ame
中文摘要 彭博周末节目在纽约直播，主持人David Gura、Christina Ruffini和Lisa Mateo回顾本周重大新闻，带来清晰解读。节目还邀请牛津移民政策与社会研究中心高级研究员Myriam Cherti参与讨论。
Berkshire Hathaway chief ends more than three-year selling streak as he ploughs a net $20bn into stocks
中文摘要 伯克希尔哈撒韦CEO格雷格·阿贝尔结束持续三年多的股票净卖出，净投入200亿美元购买股票，开始动用沃伦·巴菲特积累的巨额现金储备。
Highly-indebted companies are increasingly ditching private credit loans for cheaper capital in the bank loan market, a shift underscoring the stark realities of higher-for-longer interest rates.
中文摘要 利率持续高企背景下，高负债公司正越来越多地放弃私人信贷贷款，转向成本更低的银行贷款市场再融资，凸显私人信贷行业面临的挤压压力。
Pointed offers a strategic twist to the news quiz format, testing not just players’ knowledge of the news but also their confidence in their answers. Join Bloomberg's David Gura, Christina Ruffini and Lisa Mateo as they play and check out the quiz for yourself at Bloomberg.com (Source: Bloomberg)
中文摘要 彭博推出策略型新闻问答游戏「Pointed」，不仅测试玩家对新闻的了解程度，还考验其对答案的自信。主持人David Gura、Christina Ruffini和Lisa Mateo参与试玩。
LIV Golf is seeking fresh investments as it prepares to cut purses and move away from the huge guaranteed payments that helped it lure some of golf’s biggest stars. Author of "LIV or Let Die" Alan Shipnuck is on Bloomberg This Weekend to explain that the league’s future may depend on retaining playe
中文摘要 LIV高尔夫联赛正寻求新投资，准备削减赛事奖金并取消曾吸引顶尖球星的高额保障金。作家Alan Shipnuck表示，该联赛未来可能取决于能否获得新资金。
New Yorker staff writer Evan Osnos is on Bloomberg This Weekend and says China is taking a practical approach to AI, rapidly deploying the technology in factories and seeking to expand its use across emerging markets rather than focusing primarily on the pursuit of artificial general intelligence. (
中文摘要 《纽约客》特约撰稿人Evan Osnos表示，中国对AI采取务实路线，快速在工厂部署技术，并寻求在新兴市场扩大应用，而非专注追求通用人工智能，AI正重塑中国经济未来。
The US remains outside formal negotiations with Iran as Oman and Tehran discuss arrangements for the Strait of Hormuz, leaving the diplomatic path uncertain and raising the possibility of renewed fighting. Joining Bloomberg This Weekend, Bloomberg News White House & National Security Editor Michelle
中文摘要 美国弹药库存紧张使其对伊朗战略复杂化。美国仍未参与正式谈判，阿曼与伊朗正就霍尔木兹海峡安排进行磋商，外交路径不确定，可能再度爆发冲突。
Hunting camouflage has moved from the woods into mainstream fashion, fueled by demand for vintage workwear and rural American aesthetics. Bloomberg Weekend Senior Features Writer Madison Darbyshire is on Bloomberg This Weekend to explain how the trend has reached streetwear and luxury fashion, where
中文摘要 彭博周末专题作家Madison Darbyshire在节目中解释，狩猎迷彩服从野外走进时尚主流，受复古工装和乡村美式风格需求推动，已延伸至街头服饰等领域。
What was the thinking of the investors backing the now-canned plan, and are such proposals in the future inevitable?
中文摘要 科技投资者曾支持一项现已被搁置的世界杯相关计划，引发足球是否不受AI影响的讨论。此类提议未来是否必然重现，仍待观察。
Thousands of University of California faculty members are pushing the system to reconsider its test free admissions policy, arguing that SAT and ACT scores could help identify students prepared for college-level work as concerns grow about grade inflation and AI use. Bloomberg News Education Policy
中文摘要 数千名加州大学教师呼吁校方重新考虑「不考虑考试成绩」的录取政策，认为SAT和ACT成绩有助于识别具备大学学习能力的学生，因对成绩通胀和AI使用的担忧加剧。
Berkshire Hathaway stepped up spending under CEO Greg Abel, repurchasing about $4.5 billion of its shares while deploying billions more into investments. Bloomberg Intelligence and Senior P&C Insurance Equity Analyst Matthew Palazola is on Bloomberg This Weekend explaining to hosts David Gura and Ch
中文摘要 伯克希尔哈撒韦在CEO格雷格·阿贝尔领导下加大支出，回购约45亿美元股票，同时向多个投资项目部署数十亿美元资金，开始留下自己的治理印记。
Merrill & Bank of America Head of Market Strategy Joe Quinlan is on Bloomberg This Weekend to give his reaction to the July jobs report. Quinlan explains to hosts David Gura and Christina Ruffini that the US economy remains resilient despite the weak July jobs data, pointing to continued hiring in c
中文摘要 美银证券市场策略主管乔·昆兰表示，尽管7月就业数据疲软，但AI基础设施建设等因素支撑美国经济韧性，因此他对美股仍持看多立场。
19 回复 · 程序员 节点
12 回复 · 程序员 节点
24 回复 · 程序员 节点
21 回复 · 程序员 节点
11 回复 · 程序员 节点
86 回复 · Apple 节点
57 回复 · Apple 节点
73 回复 · Apple 节点
18 回复 · Apple 节点
11 回复 · Linux 节点
该源今日无内容。