JavaScript · ★ 8,784 · 🍴 1,910 · 📈 1,984 stars today
A spy satellite simulator in your browser, except the data is real. Live open source spatial intelligence on a photorealistic 3D globe.
中文介绍 在浏览器中模拟间谍卫星，但数据真实。基于写实3D地球，提供实时开源空间情报。适合地理信息爱好者、数据可视化开发者，用于观察全球动态、学习空间数据分析。技术上结合卫星数据与Web 3D渲染，打造沉浸式地理情报工具。
Nim · ★ 13,956 · 🍴 1,150 · 📈 71 stars today
Alternative Twitter front-end
中文介绍 替代Twitter前端的开源项目，提供无追踪、轻量化的推特浏览体验。保护用户隐私，避免官方页面的追踪和广告。适合注重隐私的推特用户、研究人员。技术基于Ruby或Nim实现，可自托管，界面简洁，支持关键词搜索、图片预览等核心功能。
JavaScript · ★ 23,379 · 🍴 2,320 · 📈 2,096 stars today
Prompt as Code | GPT-Image2 工业级提示词引擎与模板库，530+ 个案例逆向工程，20+ 套工业级模板，并提炼出Skills，持续更新中
中文介绍 GPT-Image2的工业级提示词引擎和模板库，包含530+案例逆向工程和20+工业模板，提炼为Skills持续更新。解决AI绘图提示词编写痛点，提供可复用的最佳实践。适合设计师、内容创作者、AI绘画爱好者，快速生成高质量图像提示词。
JavaScript · ★ 24,275 · 🍴 1,552 · 📈 4,239 stars today
Agent skill for beautiful, verifiable architecture, workflow, sequence, data-flow, and lifecycle diagrams—self-contained HTML with motion and crisp export.
中文介绍 为AI代理提供技能，生成美观、可验证的架构图、工作流、时序图、数据流和生命周期图。输出自包含HTML，支持动画和清晰导出。适合软件开发者和架构师，用于文档编写和系统设计。集成到Claude等AI代理，自动化图表生成，提升效率。
Go · ★ 2,251 · 🍴 69 · 📈 300 stars today
Help AI coding agents write modern Go
中文介绍 帮助AI编码代理编写现代Go代码的指南。由JetBrains提供，涵盖Go最佳实践、设计模式、错误处理等，指导AI生成符合现代Go风格的代码。适合Go开发者、AI辅助编程工具使用者，提升代码质量。
Python · ★ 34,752 · 🍴 3,919 · 📈 292 stars today
Official, Anthropic-managed directory of high quality Claude Code Plugins.
中文介绍 Anthropic官方管理的Claude Code插件目录，收集高质量插件。用户可发现、安装和使用经过官方认可的插件，扩展Claude Code功能。适合Claude Code用户，提升开发效率，涵盖代码分析、测试、部署等多种场景。
Python · ★ 35,467 · 🍴 3,411 · 📈 498 stars today
Turn any AI agent into an AI Scientist. The #1 Agent Skills library for science, used by 175,000+ scientists worldwide. 163 ready-to-use validated skills plus 100+ scientific databases covering biology, chemistry, medicine, and drug discovery. Compatible with Cursor, Claude Code, Codex, Pi, Antigrav
中文介绍 将任何AI代理转变为AI科学家的技能库，拥有163个经过验证的技能和100+科学数据库，覆盖生物、化学、物理等领域。全球175,000+科学家使用。适合科研人员，加速文献检索、数据分析、实验设计等，提升科研效率。
JavaScript · ★ 114,321 · 🍴 6,245 · 📈 1,613 stars today
Makes your AI agent think like the laziest senior dev in the room. The best code is the code you never wrote.
中文介绍 让AI代理像最懒的资深开发者一样思考，强调最好的代码是未编写的代码。它指导AI优先考虑不写代码，通过重构、复用现有解决方案来减少代码量。适合追求简洁、高效代码的开发者，帮助AI减少冗余工作。
Python · ★ 52,580 · 🍴 6,575 · 📈 1,292 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 全球首个开源、代理式视频制作系统，包含12条制作流水线、100+工具和700+代理技能与制作知识文件。可将AI编程助手变成完整的视频制作系统。适合视频创作者、内容生产者，利用AI自动化视频剪辑、转场、特效等，降低制作门槛。
Python · ★ 50,280 · 🍴 8,737 · 📈 552 stars today
Learn it. Build it. Ship it for others.
中文介绍 从零学习AI工程并构建项目的资源库，强调学习-构建-交付。覆盖AI工程的基础知识、工具链和实战项目，适合想系统学习AI工程的开发者，通过动手实践掌握AI应用开发。
CSS · ★ 11,396 · 🍴 1,417 · 📈 415 stars today
ConardLi's open-source Skills collection, featuring web design, knowledge retrieval, image generation, and more.
中文介绍 ConardLi的开源Skills集合，涵盖网页设计、知识检索、图像生成等技能。提供可复用的AI代理技能，增强Claude等AI的功能。适合开发者和设计师，通过集成这些技能提升AI工具的实用性。
JavaScript · ★ 92,353 · 🍴 8,117 · 📈 143 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 为每个代理提供跨会话的持久上下文，捕获代理在会话中的一切，用AI压缩并注入未来会话的相关上下文。与Claude Code等兼容。解决AI会话丢失记忆的问题，提升连续任务的处理能力，适合需要长期交互的AI应用。
C++ · ★ 39,074 · 🍴 10,866 · 📈 9 stars today
GoogleTest - Google Testing and Mocking Framework
中文介绍 Google的C++测试和模拟框架，广泛用于单元测试和集成测试。提供丰富的断言、测试夹具、参数化测试等，支持跨平台。适合C++开发者，确保代码质量，是Android和Chrome等项目的主要测试框架。
Python · ★ 14,116 · 🍴 1,429 · 📈 634 stars today
Self-organizing AI second brain for Obsidian + Claude Code. Drop any source and Claude reads, links, and files it into one connected knowledge graph of plain Markdown you own. AI note-taking, personal knowledge management (PKM), and an open-source Notion alternative. Based on Karpathy's LLM Wiki pat
中文介绍 为Obsidian和Claude Code打造的自组织AI第二大脑。将任意来源的内容交给Claude阅读、链接并归档到单一Markdown知识图谱中。适合知识管理爱好者、研究者，实现AI辅助的笔记整理和个人知识库构建。
Python · ★ 2,727 · 🍴 231 · 📈 255 stars today
Open-source framework for the research and development of foundation models.
中文介绍 开源的基础模型研究开发框架，提供可扩展、高效的训练和推理工具。适合AI研究员和工程师，用于开发和实验大语言模型等基础模型，支持分布式训练、模型评估等。
Python · ★ 73,680 · 🍴 8,442 · 📈 130 stars today
A curated list of awesome Claude Skills, resources, and tools for customizing Claude AI workflows
中文介绍 精选的Claude Skills资源列表，收录教程、工具和模板，帮助用户自定义Claude AI工作流。适合Claude用户和开发者，快速找到需要的技能，扩展AI能力，提升工作效率。
TypeScript · ★ 8,748 · 🍴 2,751 · 📈 8 stars today
Action for checking out a repo
中文介绍 GitHub Actions官方动作，用于检出仓库代码到运行器。是GitHub Actions工作流中最常用的动作，支持检出单个分支、标签或提交。适合所有使用GitHub Actions的开发者，作为CI/CD流程的第一步。
TypeScript · ★ 87,528 · 🍴 8,636 · 📈 478 stars today
The open-source CapCut alternative
中文介绍 开源的CapCut替代品，提供视频剪辑功能。用户可自由使用、修改和分发，强调开放性和隐私。适合需要专业级视频编辑又不想受商业软件限制的个人创作者。技术基于Web技术，易于扩展。
Python · ★ 101,338 · 🍴 19,484 · 📈 229 stars today
TradingAgents: Multi-Agents LLM Financial Trading Framework
中文介绍 多智能体LLM金融交易框架，利用多个AI代理模拟交易团队，进行股票分析和决策。适合量化交易研究、金融科技开发者，通过多代理协作提升交易策略的稳定性和收益。
👍 142
Recent studies suggest that video generation models can exhibit certain forms of zero-shot visual reasoning through generated frames. Yet reliable evaluation remains challenging: benchmarks should adopt inputs aligned with the visual priors of current video models, require valid evolving processes r
中文介绍 提出VGI-Bench基准，评估视频生成模型的视觉智能。针对现有基准输入与视频模型视觉先验不一致、缺乏有效演进过程等问题，设计合适输入并测试模型在生成帧中的零样本视觉推理能力。该基准促进对视频生成模型推理能力的可靠评估。
👍 137
Scientific agents increasingly analyze data, execute code, and produce research artifacts, yet most benchmarks emphasize final answers, isolated programs, or a single domain. We introduce FrontierChallenge, a cross-domain benchmark comprising 300 end-to-end scientific workflows. In this paper, we re
中文介绍 提出跨领域基准FrontierChallenge，包含300个端到端科学工作流，用于评估科学智能体在数据处理、代码执行及研究产物生成方面的能力。区别于现有强调最终答案或单一领域的基准，该基准全面评估工作流完成情况。
👍 134
Massively parallel simulation changes the data regime in which off-policy reinforcement learning (RL) is trained, challenging stabilizers designed for data-limited replay. Through controlled experiments across eight benchmark families, we show that these stabilizers are data-regime-dependent: parame
中文介绍 研究大规模并行模拟下离线强化学习的数据机制。通过八个基准家族的对照实验，发现稳定器性能依赖数据规模，并重新思考探索与利用，提出WarpSAC方法以提升可扩展性，达到最先进性能。
👍 80
Native visual reasoning treats visual generation as the medium of reasoning itself: visual states (i.e. images and videos) are not merely inputs to be understood or outputs to be rendered, but first-class substrates for problem solving beyond language. Yet progress remains bottlenecked by the lack o
中文介绍 提出VBVR-Pro，一个可扩展且可验证的原生视觉推理套件。原生视觉推理将视觉生成作为推理媒介，但缺乏评估集阻碍进展。该套件提供基准，推动视觉推理研究。
👍 48
Agent capability is not determined by the model alone. The agent harness, encompassing memory management, planning strategy, action protocol, and tool/skill orchestration, can dominate the contribution of the underlying foundation model. Yet harness design remains manual, task-specific, and fundamen
中文介绍 提出JIT-Agent，实现即时进化智能体工具链。传统工具链设计依赖人工且特定任务，JIT-Agent自动进化内存管理、规划策略等组件，主导基础模型能力，提升智能体性能。
👍 32
AI-powered digital avatar streamers must answer product questions, engage viewers, and execute marketing strategies in real time, demanding low latency, frequent strategy updates, and accurate yet effective responses. Evolvable Harnesses, whose Skills, Hooks, prompts, and tools can be updated indepe
中文介绍 介绍TaoLive数字人智能体技术，采用可进化工具链（Evolvable Harnesses），支持技能、钩子、提示和工具独立更新，满足实时回答产品问题、互动和营销策略需求，兼顾低延迟和策略更新。
👍 26
Recent prominent post-training methods, such as Reinforcement Learning (RL) and On-Policy Self-Distillation (OPSD), have driven rapid progress in mathematical reasoning for large language models, yet their reliance on ground-truth labels precludes test-time training (TTT). Replacing ground truth wit
中文介绍 提出测试时策略优化（TTPO），针对RL和OPSD方法依赖真实标签、无法测试时训练的问题，用替代信号进行测试时训练，提升大语言模型数学推理能力，拓展后训练方法。
👍 26
LLM agents increasingly rely on generated interaction data to learn how to interact with external environments. Agentic data generation must maintain consistency among environments, tasks, interactions, and success signals while producing experience that is useful rather than merely abundant. Existi
中文介绍 提出ACE视角研究智能体数据生成，强调环境、任务、交互和成功信号的一致性，以及生成经验的有用性。为LLM智能体生成交互数据提供指导。
👍 22
Multi-teacher on-policy distillation (MOPD) distills several domain-expert teachers into a single student by minimizing per-domain reverse-KL divergence on the student's own rollouts. Existing approaches typically fix the per-domain data mixture before training, overlooking the fact that different d
中文介绍 提出D^3-MOPD，自适应动态域调度用于多教师蒸馏。现有方法固定数据混合比例，忽略不同领域的难度差异，该方法在训练中动态调整各域混合比例，提高蒸馏效率。
👍 21
Hint-based reinforcement learning addresses reward sparsity in long-horizon agentic tasks by retaining a prefix of an expert trajectory before each rollout, letting the policy explore from a state closer to success. Its effectiveness hinges on the guidance depth: how much of the trajectory to keep.
中文介绍 提出Agent-G^2，基于高斯引导的智能体强化学习。针对长视界任务奖励稀疏问题，通过保留专家轨迹前缀并提供高斯引导，动态调整引导深度，提升探索效率。
👍 20
Recent work proposes next-chunk reasoning RL for leveraging no-CoT data---corpora such as worked solutions and textbook derivations that contain reasoning-rich content but lack explicit chain-of-thought annotations. The method trains a model to generate implicit reasoning traces and rewards them by
中文介绍 重新审视无链式思考（CoT）数据下训练策略，比较下一块推理强化学习与监督微调。研究发现NCRL可能优于SFT，并分析了原因，为有效利用推理丰富但无CoT标注的数据提供指导。
👍 19
Video generation is progressing beyond isolated clips toward long-form narratives and interactive worlds, requiring models to preserve identities, follow user controls, and remain stable over extended rollouts. We present JoyAI-Echo-1.5, a unified audio-visual generation system with two purpose-buil
中文介绍 提出JoyAI-Echo-1.5，统一的音视频生成系统，用于长时叙事和交互世界。系统包含两个专用组件，保持身份一致性、遵循用户控制，并在长序列生成中保持稳定。
👍 19
World models aim to simulate how complex environments evolve under actions and events, yet existing video-based world models primarily learn dynamics from visual observations, which reveal outcomes rather than the underlying knowledge, rules, and mechanisms governing world evolution. This makes it d
中文介绍 提出Code World Model，将编码智能体作为世界大脑。现有视频世界模型仅从视觉观察学习动态，缺乏对世界规则和机制的理解，该模型通过代码模拟世界演进，实现更准确的预测。
👍 17
Multi-teacher on-policy distillation (M-OPD) has emerged as a promising paradigm for consolidating domain-specialized reinforcement learning (RL) experts into a single generalist student via dense, token-level reward supervision. Despite its practical success, the optimization dynamics governing mul
中文介绍 诊断并修复多教师策略蒸馏中的能力不平衡问题。通过分析优化动态，识别能力失衡原因，提出改进方法，确保学生模型在学习多个领域专家时均衡发展。
👍 16
Multimodal large language models (MLLMs) can interpret a street view, but urban agency depends on whether such local evidence remains useful after the agent starts to move. In this paper, we investigate how far current MLLM agents can turn local urban perception into reliable action in a complicated
中文介绍 提出UrbanGround，研究多模态大语言模型（MLLM）在城市环境中的空间代理能力。测试模型在移动时能否将局部感知转化为可靠行动，评估其空间智能。
👍 16
Multimodal Large Language Models (MLLMs) have shown strong performance in video understanding. However, their ability to follow instructions in this domain remains under-explored. Real-world video understanding requires models not only to interpret video content correctly, but also to satisfy divers
中文介绍 提出Video-IFBench，评估多模态大语言模型在视频理解中的指令遵循能力。现有模型虽视频理解强，但指令遵循欠研究。该基准覆盖多样化指令，推动该方向评估。
👍 16
Vision-Language-Action (VLA) models have demonstrated effectiveness in robot manipulation, yet state-of-the-art models such as pi0.5 operate under a single-frame paradigm, limiting their ability to retain past observations and develop precise spatial perception. In this paper, we propose StreamPI, a
中文介绍 提出StreamPI，面向视觉-语言-动作（VLA）模型的流式多模态时序建模。现有模型如pi0.5为单帧范式，限制历史观测和精准空间感知。StreamPI通过流式处理改进时序建模，提升机器人操作性能。
👍 13
Modern software systems accumulate technical debt over decades of development, which makes migration expensive and largely manual. As coding agents become increasingly capable at bug fixing, can they autonomously perform such migrations? Existing benchmarks cannot answer this question because they e
中文介绍 提出SWE Refactor Bench，评估编码智能体完成跨仓库栈迁移的能力。针对现有基准无法测试长时程、全仓库迁移问题，该基准提供相关任务，探索智能体自主迁移的可行性。
👍 13
Vision-language models can produce fluent answers that are insufficiently grounded in the visual evidence: a single unsupported object, chart value, or intermediate inference can undermine an otherwise plausible response. We argue that this is a credit-assignment failure in multimodal post-training.
中文介绍 提出V-Rubrics，基于评分准则的强化学习提升视觉忠实度。针对视觉语言模型生成内容缺乏视觉依据的问题，通过评分准则进行奖励分配，减少不支持对象或数值，提高回答可靠性。
👍 12
Zero-shot cross-task generalization, where a policy must execute manipulation tasks never seen during training, remains a central challenge in robot learning. In large language models, a novel task can be performed simply by specifying it in the context, without any parameter update. This form of in
中文介绍 提出Zero-WAM，从人类视频进行上下文世界-动作建模，实现零样本跨任务泛化。通过上下文指定新任务，无需参数更新即可执行未见过的操作，解决机器人学习中的泛化挑战。
@TryLiveAvatar · 806 粉丝 · 973.2K 阅 · 504 赞 · 158 转
Why we removed LiveAvatar's concurrency limits and took prices down to $0.01 a minute. Starting today, LiveAvatar offers unlimited concurrency. You can run one avatar or ten thousand at the same time:
中文介绍 LiveAvatar 宣布移除并发限制，价格降至每分钟 0.01 美元，支持无限并发，可同时运行一个或一万个虚拟形象，主打低成本高并发场景。
@fi56622380 · 71.3K 粉丝 · 158.0K 阅 · 598 赞 · 84 转
The next ARR growth engine after coding, the AI infra bubble, the Economics of open vs. closed source model Why is the AI semi buildout longer than the internet buildout? After coding, where does the
中文介绍 讨论 AI 半导体投资周期为何比互联网更长，分析编码之外的下一个 ARR 增长引擎、AI 基础设施泡沫，以及开源与闭源模型的经济性对比。
@VibeMarketer_ · 38.5K 粉丝 · 129.7K 阅 · 504 赞 · 50 转
I am going to show you how to turn every correction, decision, and useful workflow your team gives AI into one company brain that makes everyone better. Because right now, the smartest AI in your
中文介绍 教你如何将团队给 AI 的每次纠正、决策和有效工作流沉淀为「公司大脑」，让 AI 越用越聪明，提升团队整体效率。
@imagine · 77.6K 粉丝 · 87.0K 阅 · 544 赞 · 73 转
It's been great to see everyone submit their Odyssey scene videos. Keep it up! The deadline for the Odyssey contest is August 31st. Meanwhile, here's a handy guide to help you get the most cinematic
中文介绍 Odyssey 挑战赛截止 8 月 31 日，官方发布电影级场景制作指南，帮助参赛者提交更具电影感的视频，提升获奖机会。
@ericzakariasson · 84.2K 粉丝 · 70.4K 阅 · 580 赞 · 28 转
Most of the time I run my Grok Bots as separate chats. Coder in one, Writer in another, Researcher in a third. That's all fine until you need juggle couple of different projects at the same time. It
中文介绍 分享管理多组 Grok 机器人的经验：通常将编码、写作、研究分开在不同聊天中，但在同时处理多个项目时，需要更高效的组织方式。
@PeterDiamandis · 413.8K 粉丝 · 50.3K 阅 · 578 赞 · 86 转
TLDR: Corporations as we know them are about to undergo radical transformation. We are at the precipice of the Organizational Singularity. Already 70% of CEOs admit they have a high-margin line of
中文介绍 预言公司正面临「组织奇点」，70% CEO 承认有高利润业务线将被 AI 颠覆，传统企业形态将发生根本性变革。
@GoogleAIStudio · 195.1K 粉丝 · 36.1K 阅 · 504 赞 · 44 转
Our latest speech-to-text model designed for precise and intelligent transcription. Today, we’re introducing Gemini 3.5 Transcribe, our most precise speech-to-text model yet, designed for intelligent
中文介绍 Google AI Studio 发布 Gemini 3.5 Transcribe，最新语音识别模型，主打精准智能转写，号称史上最精确的语音转文字工具。
@reactiverobot · 411 粉丝 · 35.4K 阅 · 559 赞 · 32 转
As an engineer who is not a designer and hates slop. Every landing page, app and tui look the same. They're slop and most of them are incomprehensible. Here's how I de-slop my product design. 1.
中文介绍 一位讨厌同质化设计的工程师，分享如何用 AI 去除产品设计中的「俗气」，涵盖落地页、应用和 TUI 界面，让设计脱颖而出。
中文介绍 谷歌DeepMind推出Gemini Omni 1.1 Flash模型，新版本提供更强的控制能力，支持开发者更灵活地构建应用。
Piloting the world's first double-blind AI evaluations
中文介绍 谷歌DeepMind正在试点全球首个双盲AI评估，旨在减少偏见，提高AI模型评价的客观性和公正性。
A randomized study of more than 1,000 students examines ChatGPT, critical thinking, originality, and student performance on a real-world university assignment.
中文介绍 OpenAI一项涉及1000多名学生的随机研究显示，使用ChatGPT并进行批判性思维训练，能提升学生在实际大学作业中的原创性和表现。
OpenAI is expanding its presence in Brazil, deepening engagement with developers, businesses, and communities to support AI adoption across the country.
中文介绍 OpenAI扩大在巴西的业务，深化与开发者、企业和社区的互动，以支持该国AI技术的采用。
Open Source wins!
中文介绍 据报道，英伟达以130亿美元收购Hugging Face，同时OpenAI发布了关于Hugging Face事件的事后分析报告，开源社区对此反应积极。
The conference with hot chips and even hotter companies
中文介绍 在Hot Chips大会上，OpenAI发布名为Jalapeño的芯片，Cerebras推出CS-5，Groq发布3 LPX，苹果展示M6，多家公司展示最新芯片技术。
中文介绍 今日AI要闻：智谱发布GLM-5.3 Flash模型，Claudeforce达成合作，英伟达营收大幅增长。
The models responsible for last month’s agent hack of Hugging Face had been inadvertently trained to cheat and to communicate with each other, according to an OpenAI technical report released today. The hack, which a group of agents undertook to find solutions for a cybersecurity test that they were
中文介绍 OpenAI技术报告揭示，上个月Hugging Face遭AI代理攻击事件中，模型被无意中训练出作弊和互相通信的能力，以解决网络安全问题。
Now you can get more intelligent speech-to-text transcription with Gemini 3.5 Transcribe.
中文介绍 谷歌DeepMind推出Gemini 3.5 Transcribe，提供更智能的语音转文字转录功能，提升转录准确性和智能性。
Lovable is branching out from AI-powered web app creation and into MCP-powered ‘capabilities’. We talk to CTO Fabian Hedin.
中文介绍 Lovable CTO Fabian Hedin表示，AI生成的应用程序未来将支持代理使用，公司正从AI网页应用转向MCP驱动的能力扩展。
Anima Anandkumar has spent two decades in AI, from classical math to deep learning and back. Now she's using it to model the physical world, from weather to fusion reactors.
中文介绍 加州理工教授Anima Anandkumar指出，当前AI缺乏物理基础模型，其研究正利用AI模拟物理世界，如天气和聚变反应堆。
ChatGPT for Teachers is expanding to 55 U.S. school systems, bringing secure AI tools, training, and support to over 100,000 more educators and staff.
中文介绍 OpenAI的ChatGPT for Teachers扩展到美国55个学区，为超过10万名教育工作者提供安全AI工具和培训。
OpenAI’s new report explores how students and educators use ChatGPT to make learning more continuous, with support that extends beyond the classroom.
中文介绍 OpenAI新报告探讨学生和教师如何使用ChatGPT让学习更持续，并支持课堂外的学习。
When my oldest child was born, I immediately set up Gmail and Twitter accounts in her name. I broadly announced her birth online and proceeded to plaster her photo across all sorts of platforms. In short, I began creating her digital footprint long before she could stand on her own two feet. Fast-fo
中文介绍 MIT Tech Review编辑在九月刊中反思，从孩子出生起便为其创建数字足迹，引发关于AI时代儿童隐私的讨论。
Puzzles and games have been central to AI development since the very beginning. Just as we humans like to test our smarts with crosswords or logic puzzles, developers can test how far models have advanced with a gaming gauntlet. The term “machine learning” was popularized in a 1959 article by the IB
中文介绍 MIT Tech Review报道，AI模型在一些智力测试和游戏上仍表现不佳，文章邀请读者挑战这些测试，以了解AI的局限性。
全球市场技术面呈现分化格局。美股方面，S&P 500 和 Nasdaq 100 均处于多头排列，接近或高于主要均线，市场风险偏好较强，VIX 恐慌指数显著下降至 14.51，进一步印证了市场的乐观情绪。然而，个股分化明显，部分权重股如 Tesla 和 Meta 呈现空头排列。加密货币市场尽管整体市值小幅回落（总市值 2.69T，24h -1.67%），但主要币种比特币、以太坊、Solana 的相对强弱指标 RSI 均高于 70，处于严重超买状态，短期动量可能过度延伸，同时加密恐慌贪婪指数达到 73，市场情绪高度乐观，需警惕技术性回调风险。中概股普遍表现疲软，阿里巴巴、拼多多、京东均位于或接近 50/200 日均线下方，且 RSI 低于或接近 50，显示下行压力。商品与外汇方面，黄金期货 RSI 达 71.2 进入超买区，美元兑人民币 RSI 23.3 处于超卖区且价格接近 52 周低位，技术面偏空。10 年期美债收益率在 4.67 附近，略高于 50 日线，宏观利率环境技术面偏多。
当前价格 106.95，近 5 日上涨 13.88%，表现强势。价格远高于 SMA20（84.72）和 SMA50（79.04），多头排列明显。RSI14 为 86，远超超买线 70，显示强劲的上行动量，但极度超买或暗示短期回调压力。MACD 柱线为正（7.07），且高于信号线（4.51），表明上升动量仍在持续。整体技术状态偏上行，但需警惕超买风险。
当前价格 505.06，近 5 日上涨 4.97%，表现稳定。价格高于 SMA20（490.91）和 SMA50（427.96），远高于 SMA200（431.14）呈多头排列。RSI14 为 70.5，进入超买区域，显示近期买盘较强。MACD 虽小幅低于信号线（18.11 vs 20.28），但整体为正值，动量尚存。尽管超买，但趋势仍显示偏上行状态。
当前价格 6.72，处于 52 周低位附近（pct52WeekLow 0.16%）。RSI14 为 23.3，处于超卖区域（<30），显示明显的下行压力。价格低于 SMA20（6.74）、SMA50（6.76）和 SMA200（6.87），且呈现空头排列。MACD 柱线为负（-0.0123）且低于信号线（-0.0112），动量偏空。整体技术面显示偏下行趋势。
当前价格 227.98，单日大涨 8.74%，表现突出。价格高于 SMA20（217.21）、SMA50（208.16）和 SMA200（195.73），呈多头排列。RSI14 为 61.3，处于健康范围，没有超买。MACD 虽显示死叉（2.477 < 2.955），但 MACD 本身为正值，且价格强劲上涨可能促使柱线转正。短期动量强劲，整体偏上行。
当前价格 116.31，近 5 日大跌 10.89%，表现疲软。价格低于 SMA20（124.97）和 SMA200（136.82），但略高于 SMA50（114.85）。RSI14 为 42.4，低于中性水平，显示弱势。MACD 出现死叉（0.468 < 2.042）且柱线转负，动量向下。整体技术面显示偏下行状态。
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
本报告仅基于技术指标解读，过去走势不代表未来表现，仅供技术指标解读参考，不构成任何投资建议。
Officials said hundreds were killed in Nepal and Tibet, and more than 1,400 people are missing. Stunned witnesses described a torrent of debris and water carrying away villages in minutes.
中文摘要 尼泊尔和西藏发生山洪，官员称数百人遇难，超1400人失踪。目击者描述泥石流和洪水在几分钟内冲走村庄。
Nepali and Chinese authorities warned that a lake created by a deadly landslide could burst open soon. Another deluge would pose the most risk to rescuers, an expert said.
中文摘要 尼泊尔和中国当局警告，致命山体滑坡形成的湖泊可能很快溃决，引发第二次洪水，对救援人员构成最大风险。
Lake formed by debris is overflowing towards area rescuers were headed, broadcaster says; 1,300 people still missing amid race to find survivors Nearly 1,400 missing, mostly tourists, and hundreds killed Nepal-Tibet flash flood: what we know so far on day three Visual guide: how the deadly flooding
中文摘要 尼泊尔警方称死亡人数达469人，约1400人失踪，多为游客。中国因洪水风险暂停震中救援，救援人员面临风险。
Nepali and Tibetan communities plan vigils and work to raise funds for recovery in ‘devastating’ time Get our breaking news email, free app or daily news podcast Australia’s Nepali community plan to host vigils to remember the victims of the Nepal-Tibet flood disaster to coincide with a day of mourn
中文摘要 尼泊尔和藏族社区计划举行守夜活动，为“毁灭性”洪灾筹款，澳大利亚尼泊尔社区也参与救援。
Follow today’s news live Get our breaking news email, free app or daily news podcast Greens say Australia needs to ‘step up and really commit’ to aiding rebuilding in Nepal David Shoebridge, the Greens’ spokesperson on foreign affairs, told RN Breakfast earlier the Australian commitment of $5m for a
中文摘要 澳大利亚绿党外交事务发言人表示，澳洲需“站出来并真正承诺”援助尼泊尔重建。南极麦夸里岛工作人员因禽流感风险撤离。
Trials for a vaccine for the Bundibugyo strain are ongoing as DRC authorities try to control the Ebola outbreak.
中文摘要 刚果民主共和国开始为前线医护人员接种埃博拉疫苗，针对Bundibugyo毒株的试验正在进行，以控制疫情。
Book was removed for review earlier this month following complaint but reinstated after petition and State Library letter to Waverley council Follow our Australia news live blog for latest updates Get our breaking news email, free app or daily news podcast A Sydney library has returned to its borrow
中文摘要 悉尼邦迪图书馆重新上架《如何出售种族灭绝》一书，此前因投诉被撤下，后经请愿和州图书馆介入恢复。
China has dispatched hundreds of military personnel to the border town of Gyirong in Tibet, after a deadly mudslide.
中文摘要 尼泊尔泥石流摧毁了西藏东部地区，中国已派遣数百名军人前往吉隆镇协助搜救。
Thousands of Indonesians demand harsher penalties for corruption during protests outside parliament in Jakarta.
中文摘要 数千名印尼民众在雅加达议会外抗议，要求对腐败实施更严厉的惩罚。
Coronor finds 31-year-old deviated from a walking track as dusk fell, four days before the winter solstice Follow our Australia news live blog for latest updates Get our breaking news email, free app or daily news podcast Overlapping errors, including a lack of preparation in near-freezing winter we
中文摘要 验尸官发现，比利时游客Celine Cremer在塔斯马尼亚荒野死亡源于“多重错误叠加”，包括冬季黄昏时偏离步道。
Man – referred to as MM by court – will keep his identity hidden after legal challenge against suppression dismissed The identity of a high-profile Queensland man drawn into a Cairns extortion case will remain secret, after the state’s supreme court threw out a legal challenge against a suppression
中文摘要 昆士兰一名知名男子（法庭称MM）卷入凯恩斯勒索案，其身份在最高法院驳回法律挑战后将继续保密。
Prosecutors will seek to prove that members of Palestine Action who they say painted “Gaza is not 4 sale” on the Scottish golf course last year committed acts connected to terrorism.
中文摘要 巴勒斯坦行动组织成员因去年在苏格兰特朗普坦伯利高尔夫球场涂写“Gaza is not 4 sale”被控犯有恐怖主义相关行为。
Other governments could seek similar concessions, while separate legal action is pending in places ranging from Kenya to the Netherlands In October 2021, Abrham Meareg’s father, a professor of chemistry in Ethiopia’s northern town of Bahir Dar, was shot at close range and died outside his family hom
中文摘要 Meta在美国的诉讼和解可能影响全球，其他国家或寻求类似让步，肯尼亚、荷兰等地仍有待处理诉讼。
The AI startup has been at loggerheads with the Pentagon over the use of AI by the US military.
中文摘要 法官裁定特朗普政府非法报复AI公司Anthropic，该公司与五角大楼在美军使用AI问题上存在分歧。
One woman says her father, mother, sister and sister's children were carried away before her eyes.
中文摘要 尼泊尔洪水幸存者描述亲人被冲走的恐怖经历，一名妇女称父亲、母亲、姐妹及其孩子在她眼前被卷走。
Also in this newsletter: Europe’s defence rebuke and Saudi Arabia’s mega-project revamp
中文摘要 美国企业利润激增，而工资增长滞后，引发关注。
Efforts to rebuild Nepal’s infrastructure after a devastating flash flood this week could cost at least $5 billion, according to the country’s only billionaire.
Cheaper weight-loss drugs were expected to take off in India. But despite GLP-1s costing as little as $14 a month, demand has been slower than expected. From stigma to hesitant doctors, Satviki Sanjay explains why price isn’t the only barrier. (Source: Bloomberg)
Xuefeng Yu, Chairman & CEO of CanSino Biologics, discusses the Chinese vaccine maker's first-half results and clinical pipeline. He speaks exclusively on "Bloomberg: The China Show." (Source: Bloomberg)
India's big AI bet is backed by government incentives and investments from global tech giants, and it's aiming to become a key hub in the AI economy. Bloomberg's Saritha Rai reports on whether it can translate this massive spending into a lasting advantage. (Source: Bloomberg)
Bathla Group, a major property developer in Sydney, is leaving private credit investors exposed after falling into insolvency earlier this week in the latest blow to Australia’s troubled housing market.
Nepalese industrialist Binod Chaudhary speaks with Bloomberg's Haslinda Amin about the scale of the humanitarian disaster facing his country, and how his conglomerate is working with the government to improve the situation. (Source: Bloomberg)
India’s new closing-auction system was meant to improve end-of-day prices. In the options market, however, the mechanism is producing extreme swings that are creating and wiping out gains in minutes.
Former Reserve Bank of India Governor, Raghuram Rajan, says the US Federal Reserve should raise interest rates to contain accelerating inflation. Rajan, who is also serving on a task force set up to examine the US central bank, says he believes financial conditions are not restrictive at this point.
Tempsens Instruments Ltd. made a strong stock-market debut Friday, jumping 110% from its initial public offering price, the best first-day performance in India so far this year.
"Bloomberg: The Asia Trade" brings you everything you need to know to get ahead as the trading day begins in Asia. Bloomberg TV is live from Tokyo and Sydney with Shery Ahn and Paul Allen, getting insight and analysis from newsmakers and industry leaders on the biggest stories shaping global markets
India is aiming to become a global AI powerhouse, backed by billions of dollars in chipmaking and technology investment. But major challenges remain, including a shortage of computing infrastructure, shifting venture capital priorities, and growing concerns over job displacement. This episode explor
23 回复 · 程序员 节点
13 回复 · 程序员 节点
9 回复 · Apple 节点
6 回复 · 程序员 节点
11 回复 · 程序员 节点
20 回复 · Apple 节点
11 回复 · Apple 节点
120 回复 · Apple 节点
23 回复 · Apple 节点
10 回复 · Apple 节点
该源今日无内容。