Swift · ★ 32,371 · 🍴 907 · 📈 2,430 stars today
A tool for creating and running Linux containers using lightweight virtual machines on a Mac. It is written in Swift, and optimized for Apple silicon.
中文介绍 苹果官方开源的 Mac 端 Linux 容器工具，基于轻量级虚拟机运行容器，使用 Swift 编写并针对 Apple Silicon 进行优化。开发者无需依赖 Docker Desktop，即可在 macOS 上原生创建和运行 Linux 容器，特别适合需要在 Mac 上构建和测试 Linux 应用的工程师。
Shell · ★ 54,679 · 🍴 5,947 · 📈 3,278 stars today
Production-grade engineering skills for AI coding agents.
中文介绍 为 AI 编程智能体提供生产级工程技能的技能库，收录了可直接用于 AI 代码 Agent 的专业能力模块。适合希望扩展 AI 编程工具能力的开发者，可用于提升 Claude Code、Cursor 等 AI 编程助手在真实工程场景中的执行质量。
Python · ★ 2,745 · 🍴 269 · 📈 426 stars today
open-source healthcare ai
中文介绍 开源医疗健康 AI 项目，致力于将人工智能技术应用于医疗健康领域。面向医疗机构、研究人员和开发者，提供开放的医疗 AI 能力支持，帮助解决医疗数据分析、辅助诊断等实际问题。
★ 16,199 · 🍴 1,684 · 📈 1,978 stars today
PM Skills Marketplace: 100+ agentic skills, commands, and plugins — from discovery to strategy, execution, launch, and growth.
中文介绍 面向产品经理的智能技能市场，汇集 100 余个 Agentic 技能、命令与插件，覆盖产品发现、战略制定、执行、发布到增长的全生命周期。适合产品经理借助 AI 提升工作效率，在各阶段快速调用专业技能模块。
Python · ★ 2,632 · 🍴 210 · 📈 319 stars today
Security scanner for AI agent skills. Detect vulnerabilities, malicious patterns, and security risks.
中文介绍 NVIDIA 推出的 AI Agent 技能安全扫描工具，可检测 AI 智能体技能包中的漏洞、恶意模式和安全风险。随着 AI Agent 技能生态快速扩张，安全人员和开发者可用它对第三方技能插件进行安全审计，防范供应链攻击。
Python · ★ 32,600 · 🍴 2,386 · 📈 661 stars today
🕵️♂️ Collect a dossier on a person by username from 3000+ sites
中文介绍 OSINT 用户名信息收集工具，可跨 3000 余个网站根据用户名自动搜集目标人物的公开档案信息。底层通过批量查询各平台的用户主页判断账号是否存在，适合网络安全研究员、渗透测试人员及隐私调查人员使用。
★ 139,863 · 🍴 34,615 · 📈 368 stars today
FULL Augment Code, Claude Code, Cluely, CodeBuddy, Comet, Cursor, Devin AI, Junie, Kiro, Leap.new, Lovable, Manus, NotionAI, Orchids.app, Perplexity, Poke, Qoder, Replit, Same.dev, Trae, Traycer AI, VSCode Agent, Warp.dev, Windsurf, Xcode, Z.ai Code, Dia & v0. (And other Open Sourced) System Prompts
中文介绍 收录了 Cursor、Claude Code、Devin、Manus、Perplexity、Replit 等数十款主流 AI 编程及助手工具的完整系统提示词与模型信息。对研究 AI 产品设计、Prompt 工程及各工具底层逻辑的开发者和研究者具有较高参考价值。
TypeScript · ★ 15,379 · 🍴 1,061 · 📈 604 stars today
Desktop app to manage markdown knowledge bases
中文介绍 用于管理 Markdown 知识库的桌面应用，支持对本地 Markdown 文件进行结构化组织和管理。适合个人知识管理、技术文档维护场景，为习惯使用 Markdown 写作的用户提供比普通文本编辑器更强的知识库管理体验。
Shell · ★ 224,804 · 🍴 19,984 · 📈 1,322 stars today
An agentic skills framework & software development methodology that works.
中文介绍 一套面向软件开发的 Agentic 技能框架与开发方法论，提供可实际落地的 AI 智能体技能组合。适合希望将 AI Agent 深度融入软件开发流程的工程团队，通过标准化技能框架提升开发效率和 AI 协作可靠性。
Go · ★ 34,151 · 🍴 1,785 · 📈 61 stars today
Fast, secure, efficient backup program
中文介绍 Go 语言编写的快速、安全、高效备份工具，支持本地磁盘、S3、SFTP 等多种后端存储。采用内容寻址存储和加密设计，支持增量备份与去重，适合个人和企业在多平台进行重要数据的可靠备份与恢复。
Shell · ★ 111,542 · 🍴 18,252 · 📈 1,599 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 一套完整的 AI 代理机构工具集，包含前端开发、社区运营、内容创作、现实校验等多种专业角色的 AI Agent，每个 Agent 都具备专属人格和工作流程。适合需要快速组建多角色 AI 团队处理复杂业务任务的开发者和创业团队。
Go · ★ 5,687 · 🍴 518 · 📈 507 stars today
Advanced DNS tunneling VPN for censorship bypass, optimized beyond DNSTT and SlipStream with low-overhead ARQ, resolver load balancing, high packet-loss stability and speed.
中文介绍 基于 DNS 隧道技术的高级 VPN 工具，在 DNSTT 和 SlipStream 基础上进行优化，具备低开销 ARQ 重传机制、DNS 解析器负载均衡及高丢包环境下的稳定性。适合网络受限环境下需要突破封锁的用户和开发者。
Ruby · ★ 30,353 · 🍴 7,506 · 📈 67 stars today
Open-source live-chat, email support, omni-channel desk. An alternative to Intercom, Zendesk, Salesforce Service Cloud etc. 🔥💬
中文介绍 开源全渠道客户支持平台，集在线客服、邮件支持、全渠道工单管理于一体，可作为 Intercom、Zendesk 等商业产品的自托管替代方案。支持 WhatsApp、Twitter、Email 等多渠道接入，适合中小企业和技术团队自建客服系统。
Go · ★ 1,634 · 🍴 173 · 📈 114 stars today
Local-first session intelligence and analytics for coding agents, supporting Claude Code, Codex, and more than 20 other agents. Also: 100x faster replacement for ccusage!
中文介绍 本地优先的 AI 编程 Agent 会话智能分析工具，支持 Claude Code、Codex 等 20 余种 Agent，可统计会话数据、分析使用模式。同时号称比 ccusage 快 100 倍，适合希望深入了解自身 AI 编程助手使用情况的开发者。
★ 7,938 · 🍴 2,413 · 📈 89 stars today
张雪峰.skill — 张雪峰的认知操作系统。高考志愿/考研/职业规划的实战思维框架。由女娲.skill生成。
中文介绍 以张雪峰为人设构建的 AI 技能模块，封装其在高考志愿填报、考研和职业规划领域的实战思维框架，由女娲.skill 生成。适合高中生、家长及需要升学和职业规划指导的用户，以对话形式获取类张雪峰式的实用建议。
Roff · ★ 73,918 · 🍴 16,537 · 📈 88 stars today
所有小初高、大学PDF教材。
中文介绍 收录了小学、初中、高中及大学各阶段的 PDF 版中国教材，资源覆盖全面。适合学生自学、教师备课、家长辅导以及需要查阅各版本国内教材的用户，提供免费的教材电子版获取渠道。
Python · ★ 1,288 · 🍴 158 · 📈 199 stars today
SIA is a Self Improving AI framework to autonomously improve the performance of any AI system (Model / Agent) on a benchmark task.
中文介绍 SIA 是一个自我优化 AI 框架，能在指定基准任务上自主提升任意 AI 系统（模型或 Agent）的性能。通过自动化迭代改进机制减少人工干预，适合 AI 研究人员和工程师用于持续优化模型在特定任务上的表现。
TypeScript · ★ 37,308 · 🍴 8,697 · 📈 53 stars today
Mattermost is an open source platform for secure collaboration across the entire software development lifecycle..
中文介绍 开源企业级安全协作平台，覆盖软件开发全生命周期的团队沟通需求，提供消息、文件共享、工作流自动化等功能。支持私有化部署，适合对数据安全有严格要求的软件研发团队替代 Slack 等 SaaS 协作工具。
Kotlin · ★ 46,750 · 🍴 8,004 · 📈 161 stars today
翻墙-科学上网
中文介绍 科学上网与翻墙工具资源汇总项目，收录多种主流翻墙方案、工具配置教程及相关资源链接。面向需要访问境外互联网资源的中文用户，提供 VPN、代理等多种网络自由访问方式的综合指引。
👍 3
There are two main Parameter-Efficient Fine-Tuning (PEFT) techniques for Large Language Models (LLMs). While Low-Rank Adaptation (LoRA) introduces additional weights between the LLM layers, Soft Prompting introduces additional fine-tuning-specific raw tokens to an LLM input. However, both require mo
👍 1
Modern LLM training pipelines increasingly rely on other models to generate data, filter corpora, judge outputs, and guide development decisions. These dependencies are recursive: a model may depend on an upstream artifact whose own dependencies are documented only in separate releases and artifacts
👍 15
Vision-language models (VLMs) project images into hundreds to thousands of visual tokens, making decoder inference expensive in both attention computation and KV-cache memory. Existing visual-token reduction methods largely follow a rank-and-remove paradigm: they score visual tokens, keep a compact
👍 2
Large language models (LLMs) are widely used to tackle complex tasks with autonomous workflows. Recently, reusable natural language skills have emerged as a popular paradigm to inject procedural knowledge into LLM applications. Since popular skills are often invoked repeatedly, placing their full te
👍 17
Large Language Models (LLMs) are increasingly used for code generation, raising concerns that they may be misused to produce malicious code. Meanwhile, Grammar-Constrained Decoding (GCD) has been widely adopted to improve the reliability of LLM-generated code by enforcing syntactic validity. In this
👍 3
Remaining Useful Life (RUL) prediction is essential for industrial predictive maintenance, yet many learning-based approaches rely on extensive feature engineering or large labeled datasets to train task-specific sequence models. In this work, we introduce a lightweight learning approach, in which w
👍 28
Spatial reasoning from egocentric videos is inherently challenging because the observable evidence is constrained by the camera trajectory. Existing methods rely on single-turn inference, forcing models to resolve geometric ambiguity through semantic priors rather than verifiable evidence. We argue
👍 55
General-purpose agents such as OpenClaw are increasingly used as autonomous tool users, but their coding ability is difficult to measure under SWE-bench: a generic agent does not by itself satisfy the clean Docker workspace, patch, and prediction contract required for scoring. We introduce Claw-SWE-
👍 68
Scientific progress depends on a repeated loop of exploration, experimentation, and abstraction. Researchers test candidate directions, interpret the evidence, and carry the resulting lessons into later attempts. We study how an AI agent can run this loop autonomously over long horizons. We introduc
👍 6
Reinforcement Learning (RL) with verifiable environments has emerged as a powerful approach for enhancing the reasoning capabilities of Large Language Models (LLMs). While prior research demonstrates that scaling environment quantity improves RL performance, existing manual or individual constructio
👍 56
Environments serve as interactive systems for large language model (LLM) based agents across diverse scenarios and play a crucial role in driving the continual evolution of model capabilities. Despite this importance, existing work lacks a systematic categorization and deep analysis. This paper syst
👍 17
Recent progress in foundation models has shifted toward agentic behavior involving multi-step reasoning and tool use. However, open-source efforts largely focus on text-dominant settings, leaving long-horizon multimodal tasks underexplored. This gap is evident in video tasks requiring sustained temp
👍 6
Pretrained video generators are promising visual world models that exhibit emergent task-solving abilities; however, their reliance on detailed textual descriptions limits their direct use for planning and decision-making. Existing approaches either outsource this reasoning to language or vision-lan
👍 1
Understanding and predicting how social beliefs evolve in response to events -- from policy changes to scientific breakthroughs -- remains a fundamental challenge in social science. Given LLMs' commonsense knowledge and social intelligence, we ask: Can LLMs model the dynamics of social beliefs follo
👍 15
Reinforcement learning with verifiable rewards (RLVR) is a promising approach for enhancing reasoning and agentic behavior in large language models. However, rollout-intensive policy optimization is often limited by insufficient reward contrast, arising when overly simple or complex prompts generate
👍 1
Large language models (LLMs) are on the rise for accelerating scientific discovery, most recently in advanced tasks such as generating valid scientific hypotheses. Yet in many discovery settings, the goal is not to identify a single best hypothesis since validation can be noisy and expensive, and sc
👍 18
Combinatorics is central to Olympiad-level mathematical problem solving, requiring deep discrete reasoning, creative constructions, and rigorous structural insight. Recent evidence suggests that even today's strongest frontier models remain uneven on Olympiad combinatorics, revealing a gap in creati
👍 8
We introduce Embodied-R1.5, a unified Embodied Foundation Model (EFM) that integrates comprehensive embodied reasoning capabilities, spanning embodied cognition, task planning, correction, and pointing, within a single architecture toward general physical intelligence. Leveraging three automated dat
👍 5
Diffusion models have consistently driven progress in text-to-image generation. However, it is challenging to attribute recent progress to specific modeling and data choices: state-of-the-art open-weight models provide limited ablations, and do not disclose their training data and full training deta
👍 1
As recommender systems transition toward agentic, multi-turn conversational interfaces, evaluation paradigms have struggled to keep pace. Current benchmarks often rely on "LLM-as-a-judge" evaluations, which introduce subjectivity, high costs and inconsistency. We present τ-Rec, a benchmark for agent
👍 1
Brain Magnetic Resonance Imaging (MRI) plays a central role in studying neurological development, aging, and diseases. One key application is Brain Age Prediction (BAP), which estimates an individual's biological brain age from MRI data. Effective BAP models require large, diverse, and age-balanced
👍 51
Reward models are central to text-to-image post-training, but visual preference is subjective and better represented as a distribution over rubric scores than as a deterministic scalar. Existing scalar, score-token, and pairwise reward models over-compress uncertainty and fine-grained score differen
👍 4
Agent skills provide a lightweight mechanism for extending general-purpose agents, but their open format exposes them to skill-poisoning attacks. A practically dangerous injection must stay invisible: if executing the payload derails the user's legitimate task, the resulting failure signal invites i
👍 4
Computer-use agents (CUAs) rely on visual observations of graphical user interfaces, where each screenshot is encoded into a large number of visual tokens. As interaction trajectories grow, the token cost increases rapidly, limiting the amount of history that can be incorporated under fixed context
👍 3
Many modern vision-language models (VLMs) build on autoregressive decoding of discrete tokens. While text-based output interfaces enable scalable pretraining and strong zero-shot generalization across diverse tasks, they are poorly suited for problems that require precise continuous outputs, such as
👍 1
Sparse attention reduces compute and memory bandwidth for long-context LLM inference. However, two key challenges remain: (1) KV cache capacity still grows with sequence length, and offloading to CPU memory introduces a PCIe transfer bottleneck; (2) the sparse selection step itself retains O(T^2) co
👍 1
Curating training data is among the most consequential yet labor-intensive parts of modern AI development: practitioners iteratively propose, implement, evaluate, and revise data policies against noisy benchmark feedback. We ask whether generalist coding agents can automate this data-curation loop.
👍 3
Prior work has shown that instruction-tuned large language models (LLMs) are less well calibrated than their base pre-trained counterparts. However, little is known about the frequently used chat template's effect on the calibration of conversational LLMs. In this work, we investigate the mechanisms
👍 9
Autonomous LLM training is often framed as recipe search, which leaves the training harness largely static. This limitation sharpens in agentic RL, where shifting bottlenecks and scalar rewards mask diverse failure modes. We introduce EvoTrainer, an autonomous training framework that co-evolves LLM
👍 1
Post-training for reasoning models typically combines supervised fine-tuning with reinforcement learning from verifiable rewards, most commonly with GRPO. However, this algorithm suffers from sparse rewards, limited exploration, and mode collapse. Building upon recent works on self-distillation, we
@sairahul1 · 113.0K 粉丝 · 852.6K 阅 · 600 赞 · 79 转
Peter Steinberger, creator of OpenClaw, who now works with OpenAI. Yesterday he posted this: "You shouldn't be prompting coding agents anymore. You should be designing loops that prompt your agents."
中文介绍 OpenClaw 创始人、现 OpenAI 成员 Peter Steinberger 提出新范式：2026 年 AI 工程师应该停止手动给编程 agent 写 prompt，转而设计「循环（loop）」来驱动 agent 自动运行。这一观点正在 AI 工程圈引发讨论，标志着从「提示工程」向「循环工程」的范式转变。
@dhaber · 50.0K 粉丝 · 497.3K 阅 · 500 赞 · 57 转
One of the biggest ways that AI is transforming work (and also one of the most taboo subjects inside companies at the moment) is that most work discussions are being recorded now by default. This
中文介绍 AI 正在悄然改变职场生态：工作讨论默认被录音已成常态，但这在公司内部仍是高度敏感的话题。作者探讨这一趋势对工作方式的深远影响——当所有对话都被记录，组织文化、责任追究与信任关系都将面临根本性重塑。
@saranormous · 143.5K 粉丝 · 194.8K 阅 · 614 赞 · 40 转
The mid-2026 investor's version of AI psychosis is a despair that nothing is investable, that we should put all our money into Anthropic and Nvidia and go home. I have never felt it. I have been sure
中文介绍 作者反驳 2026 年投资圈的「AI 悲观论」——部分投资人认为除 Anthropic 和 Nvidia 外已无可投标的，陷入集体焦虑。作者表示自己从未持此看法，坚信 AI 赛道仍有大量值得押注的机会，并分享对当前投资格局的独立判断。
@0x_rody · 1.7K 粉丝 · 193.2K 阅 · 513 赞 · 72 转
Claude writes your code, hands it over, and 3 tests are failing. You paste the errors back, it fixes one thing, breaks another, and you spend the evening as a messenger between Claude and your
中文介绍 针对 Claude Code 编程时「修一处破另一处」的死循环痛点，作者分享如何搭建自改进循环（self-improving loop）：让 Claude 在测试失败后自动获取错误信息、自主修复，无需人工反复粘贴错误日志充当「传话员」，大幅降低调试摩擦。帖子含具体配置方案。
@MatthewBerman · 121.3K 粉丝 · 108.0K 阅 · 661 赞 · 26 转
tl;dr I've been testing Fable (Mythos) for the past week and it feels unlike any other model I've used. It feels, and is priced, like a next-generation model. It also has some real quirks. The Good
中文介绍 Matthew Berman 测评新模型 Fable（Mythos 系列）一周后的深度总结：该模型使用体验与定价均接近「下一代」水准，感受与其他模型明显不同，但也存在若干明显怪癖。帖子涵盖优缺点分析，适合关注前沿模型动态的开发者参考。
@Kimi_Moonshot · 172.7K 粉丝 · 106.6K 阅 · 500 赞 · 61 转
Our predictions will probably be wrong. But the World Cup offers a rare, public, verifiable, and constantly evolving real-world setting. Through this initiative, we hope to place analysis,
中文介绍 Kimi 官方宣布将预测 2026 世界杯全部 104 场比赛结果，并认为德国队被低估。此举将世界杯作为公开、可验证、持续演进的真实世界测试场，旨在展示 AI 实时分析能力，同时坦诚承认预测结果大概率存在偏差。
@0xCodez · 5.3K 粉丝 · 97.8K 阅 · 510 赞 · 80 转
Most developers still prompt their coding agents by hand. They type, they wait, they read the diff, they type again. 9out of 10 builders have never written a single loop that prompts the agent for
中文介绍 作者整理从「手动提示者」进阶为「循环设计师」的 14 步路线图。当前 90% 的开发者仍在手动与编程 agent 交互，从未写过一个自动驱动 agent 的循环。帖子系统拆解循环工程的核心概念与实践路径，面向想提升 AI 开发效率的工程师。
@RLanceMartin · 30.4K 粉丝 · 84.7K 阅 · 660 赞 · 50 转
Mythos-class models like Claude Fable 5 have changed the way many of us work at Anthropic. I want to share two tips for getting the most out of this class of models. Self-correction loops There’s been
中文介绍 Anthropic 员工分享 Claude Fable 5（Mythos 级模型）的两条使用技巧：一是充分利用自我修正循环（self-correction loop），二是另一条未在预览中披露的建议。作者表示该级别模型已改变 Anthropic 内部团队的实际工作方式。
@RayDalio · 2.2M 粉丝 · 72.6K 阅 · 515 赞 · 93 转
What is the best approach to being effectively intelligent now that human intelligence and artificial intelligence are merging? Because I have been building computerized investment decision-making
中文介绍 瑞·达利欧探讨人类智能与人工智能融合时代的「有效思考」方法论。凭借数十年计算机化投资决策系统的构建经验，他认为原则性思维与 AI 必须结合使用，并分享如何在 AI 时代保持独立判断与系统化决策能力。
@nifinet · 10.2K 粉丝 · 60.9K 阅 · 522 赞 · 54 转
When a team says they want AI for growth, they usually mean a faster send. An agent that fires the same template at a longer list, day and night. That is the cheap half of the job, and it stopped
中文介绍 作者分享用 Claude Code 构建 AI 驱动的 GTM（市场进入）增长大脑的思路。批评「AI 用于增长」常被简化为「更快发送同一套模板」，指出真正有价值的是能理解客户意图、动态生成策略的 agent，而非简单自动化，并提供具体搭建框架。
@polynoamial · 129.2K 粉丝 · 43.7K 阅 · 707 赞 · 80 转
tl;dr: As LLMs become more capable, benchmark performance is increasingly a function of test-time compute. In fact, we likely don't know what the capability ceiling is for modern LLMs because it's too
中文介绍 研究者指出：随着 LLM 能力提升，基准测试分数越来越取决于推理时算力（test-time compute）的投入，而非模型本身的能力上限。当前我们可能严重低估了现有 LLM 的真实能力天花板，因为极限算力下的性能尚未被充分探索。
@GoogleAIStudio · 176.3K 粉丝 · 32.2K 阅 · 517 赞 · 55 转
Twenty years ago, translation at Google began as one of our pioneering machine learning experiments to turn the science of language into the magic of human connection. That experiment has come a long
中文介绍 Google 推出基于 Gemini 3.5 的实时语音翻译功能 Live Translate，支持流畅、自然的语音跨语言转换。Google 将此定位为其 20 年机器翻译探索的里程碑式成果，标志着 AI 实时语音翻译从实验室走向实用化的重要节点。
@HYPERPEPS · 3.4K 粉丝 · 4.4K 阅 · 720 赞 · 79 转
You spend twenty minutes writing a brief. Then you open a second tab to turn it into code. A third for design. A fourth for the copy. Then you copy the output from tab four back into tab one, because
中文介绍 HYPERPEPS 介绍为 Hyperliquid 生态构建的 AI 层产品，解决多标签页切换、手动复制输出的割裂工作流痛点——从撰写简报到生成代码、设计、文案，一站完成。定位为 Hyperliquid 此前缺失的 AI 整合层，面向链上 DeFi 用户提升操作效率。
Google DeepMind is funding research into the potential dangers of situations where millions of different AI agents interact with each other online. According to Rohin Shah, who directs the company’s AGI safety and alignment research, the mass-market arrival of agents that can carry out tasks without
中文介绍 谷歌DeepMind正资助研究数百万AI智能体相互交互时可能引发的安全风险。该公司AGI安全与对齐研究主管Rohin Shah表示，AI智能体大规模进入消费市场后，彼此之间的复杂交互可能带来难以预测的危险，DeepMind已将多智能体安全问题列为重要研究方向。
a quiet day lets us reflect on a great essay
中文介绍 Latent Space发布AI新闻简报，探讨开放模型、模型实验室与智能体实验室之间的竞争格局，以及AI领域中哪些能力目前仍难以通过训练获得等议题，并收录了Sarah Guo撰写的相关评论文章。
OpenAI plans to acquire Ona to expand Codex with secure, persistent cloud environments, enabling long-running AI agents across enterprise workflows.
中文介绍 OpenAI宣布计划收购Ona公司，旨在通过整合安全、持久的云端环境来扩展Codex的功能，使AI智能体能够在企业工作流程中执行长周期任务，进一步推动企业级AI自动化部署。
OpenAI supports the EU Code of Practice on AI content transparency, advancing provenance standards and tools to help people understand AI-generated content.
中文介绍 OpenAI宣布支持欧盟《AI内容透明度行为准则》，致力于推进AI生成内容的溯源标准与相关工具建设，帮助用户识别和理解AI生成内容，配合欧盟构建可信赖AI生态系统的监管努力。
Discover how astrophysicist Chi-kwan Chan uses Codex to build black hole simulations, helping scientists study extreme physics and test Einstein’s theory of general relativity.
中文介绍 天体物理学家Chi-kwan Chan介绍其使用OpenAI Codex辅助构建黑洞模拟程序的实践，该工具帮助科学家研究极端物理现象并验证爱因斯坦广义相对论，展示了AI编程工具在前沿科学研究中的应用价值。
Learn how BBVA scaled ChatGPT Enterprise to 100,000 employees and partnered with OpenAI to accelerate AI-powered banking transformation worldwide.
中文介绍 西班牙对外银行BBVA宣布与OpenAI深化合作，已将ChatGPT Enterprise推广至全行10万名员工，双方将共同加速AI驱动的全球银行业务转型，将人工智能置于银行核心战略地位。
中文介绍 Hugging Face发布PyTorch性能分析系列教程第二篇，内容涵盖从基础nn.Linear层到融合MLP（多层感知机）的优化过程，介绍如何通过算子融合技术提升模型推理效率，适合深度学习工程师参考。
中文介绍 TLDR AI 2026年6月11日简报涵盖：Anthropic CEO Dario Amodei的AI政策动态、谷歌DeepMind发布DiffusionGemma文本生成加速模型，以及WhatsApp计划开放机器人账号接入等多条AI行业资讯。
Access OpenAI models and Codex through Oracle Cloud, using existing commitments to build and deploy AI with enterprise security and governance.
中文介绍 OpenAI与甲骨文达成合作，企业客户可通过现有Oracle云服务承诺额度直接访问OpenAI模型及Codex，在具备企业级安全与合规治理框架的环境下构建和部署AI应用，降低企业接入门槛。
中文介绍 谷歌DeepMind发布DiffusionGemma模型，采用扩散式生成架构实现文本生成速度提升约4倍，与传统自回归语言模型相比在推理效率上有显著突破，为大规模文本生成应用提供更快速的解决方案。
A new report from OpenAI details PRC-linked influence operations using AI to target U.S. tech debates, data center narratives, tariffs, and false claims about ChatGPT.
中文介绍 OpenAI发布新报告，揭露与中国政府相关联的网络影响行动利用AI工具干预美国科技政策讨论，目标涵盖数据中心建设叙事、关税政策辩论，并散布关于ChatGPT的虚假信息，意图影响美国AI政策走向。
Google DeepMind and partners announce a $10M funding call for multi-agent safety research.
中文介绍 谷歌DeepMind联合合作伙伴宣布设立总额1000万美元的多智能体AI安全研究专项资金，面向外部研究机构公开征集研究提案，重点资助多个AI智能体相互交互场景下的安全风险识别与防控研究。
The much anticipated launch of the Mythos-class model was marred by some controversial usage policies
中文介绍 Anthropic发布备受期待的Claude Fable 5旗舰级模型（Mythos级别），但此次发布因部分争议性使用条款而引发外界批评。模型性能表现被普遍看好，惟相关政策规定引发社区讨论，为此次发布蒙上阴影。
See how LSEG uses OpenAI to scale trusted AI across its global business, accelerating insights, shrinking release cycles, and empowering 4,000 employees.
中文介绍 伦敦证券交易所集团（LSEG）介绍其与OpenAI合作推进AI规模化应用的实践：通过AI工具加速业务洞察获取、缩短产品发布周期，目前已有4000名员工使用相关AI能力，助力全球业务智能化升级。
中文介绍 TLDR AI 2026年6月10日简报涵盖：Anthropic发布Claude Fable 5新模型、谷歌推出Gemini 3.5实时翻译功能，以及测试时算力扩展（scaling test time compute）相关研究进展等多条AI行业重要动态。
当前各资产技术面呈明显分化格局。美股方面，SPY 与 QQQ 维持多头排列（SMA50 与 SMA200 均在价格下方），但两者 MACD 均低于信号线，短期动量趋缓；SPY 的 RSI14 为 50.3，QQQ 为 53.6，均处于中性区间。个股层面，META 与 MSFT 已形成空头排列并触发 MACD 死叉，META 的 RSI14 低至 35.0，MSFT 为 36.8，空头结构相对显著；TSLA 单日涨幅达 4.6%，但 MACD 仍处负值区间，5 日仍累计下跌 4.61%，方向信号混杂。加密资产方面，BTC、ETH、SOL 三者均处于完整空头排列，RSI14 分别为 32.6、30.9、34.2，逼近超卖区间，加密恐慌贪婪指数录得 12（极度恐慌），加密总市值为 2.26T 美元，24 小时变动 -2.53%，BTC 主导率 56.4%，情绪与技术结构双重承压。中概股板块整体技术面偏弱，BABA 的 RSI14 已跌至 29.4 进入超卖，BABA、PDD、JD、腾讯均处于空头排列，5 日跌幅普遍在 3%-10% 之间。商品与外汇方面，黄金期货（GC=F）RSI14 报 35.3，距 SMA20（4,452）偏离较大；WTI 原油单日跌 4.11%，5 日跌 7.21%；美元指数 DXY 表现突出，多头排列成立，RSI14 达 59.3，为宏观类资产中技术结构最为强势的标的。10 年期美债收益率（^TNX）当前报 4.46%，维持多头排列，但 MACD 已低于信号线，短期动量偏弱。VIX 当日大幅回落 12.51%，但 5 日仍上涨 26.23%，短期波动率结构尚未完全平息。
META 当前价 568.43，处于空头排列：SMA20（606.78）、SMA50（622.07）、SMA200（659.02）均在价格上方且呈递减排列。RSI14 报 35.0，逼近超卖区间。MACD 为 -11.38，信号线为 -6.76，柱状图持续走负，已触发死叉信号。5 日跌幅达 9.42%，距 52 周高点回撤 28.61%，动量结构偏弱，技术面呈全面空头形态。
BABA 当前价 112.69，RSI14 仅报 29.4，已进入超卖区间，系本次扫描中读数最低的标的之一。三条均线全面空头排列：SMA20（127.22）、SMA50（130.55）、SMA200（149.63）均大幅高于现价。MACD 为 -4.54，信号线 -2.95，负向扩散。5 日跌幅 10.53%，距 52 周高点回撤 41.51%，超卖状态下动量仍未见收敛迹象。
QQQ 当前价 717.12，日内涨幅 3.38%，为美股组中单日表现最强标的。SMA50（679.07）与 SMA200（624.64）均在价格下方，多头排列成立。然而 MACD（8.69）低于信号线（14.93），显示短期动量趋缓；RSI14 报 53.6，处于中性区间。5 日跌幅 -3.17% 与单日强反弹并存，日线级别存在方向分歧，技术信号尚未明确。
BTC 当前价 63,491，三条均线完全空头排列：SMA20（68,795）、SMA50（74,712）、SMA200（78,008）均显著高于现价。RSI14 报 32.6，接近超卖。MACD 为 -3,994，信号线 -3,514，负向扩散持续。结合加密恐慌贪婪指数仅为 12（极度恐慌）及加密总市值 24 小时跌幅 2.53%，技术与情绪面共同指向空头结构，距 52 周高点回撤 49.69%。
DXY 当前价 99.77，SMA20（99.37）、SMA50（98.89）、SMA200（98.64）均在价格下方，形成多头排列。RSI14 报 59.3，为本次全部标的中偏高读数，动量向上。MACD（0.31）高于信号线（0.25），金叉形态成立。现价距 52 周高点仅偏离 0.86%，处于强势区间，技术结构整体偏多。
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所得技术指标进行客观描述，过去走势不代表未来表现，所有技术信号均存在失效可能。本内容仅供技术指标解读参考，不构成任何投资建议、买卖推荐或收益承诺，读者须自行承担相应决策风险。
Claiming there was progress in peace negotiations, President Trump said he had canceled the next wave of planned attacks after two days of U.S. airstrikes.
中文摘要 特朗普声称美伊和平谈判取得进展，宣布在美国对伊朗发动两天空袭后取消下一轮计划中的打击行动，并再次表示和平协议即将达成。
Revolutionary Guard cited as saying Iranian forces stopped tanker from crossing waterway without coordination; US president says signing could happen in Europe this weekend Full report: Trump says US and Iran on verge of signing peace agreement Three Indian seafarers were killed in a US attack on an
中文摘要 据报道，伊朗革命卫队阻止一艘油轮在未经协调的情况下通过霍尔木兹海峡。与此同时，特朗普称美伊和平协议"很快"将在欧洲签署。事件中另有三名印度海员遇难。
One Nation leader says Rinehart has been ‘very beneficial’ to the formulation of the party’s policies. Follow today’s news live Get our breaking news email, free app or daily news podcast Teens who use social media two hours daily at higher risk of depressive symptoms, study finds Teenagers who spen
中文摘要 澳大利亚一国党领袖波林·汉森表示，矿业富豪吉娜·莱因哈特是她的"朋友"，并对该党政策制定"大有裨益"。此外，一项研究显示每日使用社交媒体两小时以上的青少年出现抑郁症状的风险更高。
Mexico's two goal scorers were overcome with emotion as they took their team to a win in the World Cup's opening game.
中文摘要 在2026年世界杯揭幕战中，墨西哥队凭借胡利安·基尼奥内斯和劳尔·希门尼斯的进球赢得胜利，两名进球球员赛后情绪激动。
US President Donald Trump claims Washington and Tehran have reached a ‘great settlement’ and are finalising documents
中文摘要 美国总统特朗普宣布，美国与伊朗已达成"重大和解"，双方目前正在完成相关协议文件的最终确认工作。
SpaceX’s IPO could challenge pension funds as concerns grow over its valuation and governance structure under Musk.
中文摘要 SpaceX拟进行IPO，估值达1.8万亿美元。分析人士指出，由于其估值水平及马斯克主导下的公司治理结构存在隐患，该IPO对部分养老基金等机构投资者而言可能"极为不利"。
Mr. Trump said that Iran was close to signing a peace deal. So far, weeks of talks have failed to produce an agreement.
中文摘要 特朗普最新转向，撤回对伊朗发动新一轮打击并扩大战争规模的威胁，声称伊朗即将签署和平协议。但此前数周谈判均未能达成任何协议。
The officer was shot on Thursday while investigating the firing of a gun outside the U.S. consulate in March, the authorities said.
中文摘要 加拿大多伦多一名警察在调查今年3月美国领事馆外枪击事件时，于周四遭枪击身亡，当局已就此展开调查。
In addition to concerns about their safety in the event of another all-out war, many Iranians worry about the country’s economy further collapsing if the conflict remains in limbo.
中文摘要 伊朗民众在恐惧与宽慰之间煎熬，既担忧新一轮全面战争带来的人身安全威胁，也忧虑冲突持续僵局将导致本已脆弱的伊朗经济进一步崩溃，多数人渴望战争尽快结束。
Group of lawmakers call on Trump administration to facilitate medical evacuations out of Gaza amid dearth of services.
中文摘要 多名美国国会议员联名致函，敦促特朗普政府向以色列施压，允许加沙癌症患者转往境外接受治疗，理由是加沙地区医疗资源严重匮乏。
Woman, 58, was last heard from Monday afternoon and her car was found the next day Follow our Australia news live blog for latest updates Get our breaking news email, free app or daily news podcast A man has been charged over the alleged kidnapping and murder of a woman whose body is yet to be found
中文摘要 澳大利亚悉尼一名58岁女性自周一下午失联，次日其车辆被发现，本人下落不明。警方已就该女性遭绑架及谋杀案对一名男子提起诉讼，但其遗体尚未寻获。
Murder of Melissa and Mark Hortman by man disguised as police officer prompted concerns about political violence in US.
中文摘要 一名男子对在美国明尼苏达州杀害民主党高层人士梅利莎·霍特曼及其丈夫马克·霍特曼一事认罪。该男子案发时伪装成警察，此案引发外界对美国政治暴力问题的广泛关注。
The deal to ship 250,000 barrels of fuel to Cuba could have eased an energy crisis. But the Trump administration says Vanguard Energy lacks the authorization to proceed.
中文摘要 特朗普政府以缺乏授权为由，阻止佛罗里达州先锋能源公司向古巴运送25万桶燃油的交易，该交易原本有望缓解古巴当前的能源危机。
Elon Musk and people in his orbit are about to get much, much richer.
中文摘要 随着SpaceX IPO进程推进，埃隆·马斯克及其身边人士的财富将大幅增加。
Iranian leadership has not confirmed claim, but US president says planned strikes on Iran cancelled Middle East crisis – live updates Donald Trump claimed on Thursday that the US and Iran are on the verge of signing a peace agreement and announced that he will cancel fresh missile strikes. His comme
中文摘要 特朗普宣称美国与伊朗即将签署和平协议，并表示已取消对伊朗新一轮导弹打击计划。但伊朗领导层尚未证实上述说法。
Gold held its biggest gain since March after President Donald Trump said the US could sign a deal with Iran over the weekend to end the war that’s rattled global markets and stoked inflation.
中文摘要 特朗普表示美国可能于本周末与伊朗签署和平协议，终结扰动全球市场、推升通胀的中东冲突。受此消息提振，黄金维持3月以来最大单日涨幅，市场对地缘风险溢价的押注有所降温。
The public sale is also expected to make Elon Musk the world's first trillionaire.
中文摘要 SpaceX完成历史规模最大IPO前的融资，募集资金750亿美元，股票发行定价为每股135美元，创下全球IPO纪录。此次上市预计将使埃隆·马斯克成为全球首位身家超过万亿美元的富豪。
MediaTek Inc. shares are poised for their best quarter on record, as investors bet a shift into artificial intelligence chips can help it shed the overhang of its struggling older-tech business.
中文摘要 联发科股价有望创下史上最佳季度表现，投资者押注该公司加速向人工智能芯片领域转型，有望摆脱传统业务低迷带来的估值压制，从行业落后者转变为AI芯片市场的有力竞争者。
From trade wars to soaring ticket prices, the 2026 World Cup is unlike any before it. Faisal Islam explores what this tournament reveals about our changing global economy.
中文摘要 2026年世界杯在贸易战背景下举办，票价大幅攀升，被认为是经济环境最复杂的一届杯赛。BBC经济编辑法伊萨尔·伊斯兰分析本届赛事如何折射出全球经济格局的深刻变化。
Japanese equities rallied after US President Donald Trump signaled the US is close to signing a deal with Iran, fueling expectations that the Middle East conflict is nearing an end.
中文摘要 受特朗普表示美伊接近达成协议、中东冲突或将外交收场的消息提振，日本股市全线上涨，市场情绪明显改善，投资者对地缘政治风险缓和预期升温。
Agave plants grow wild in India and new distillers are using them to create a spirits industry.
中文摘要 印度野生龙舌兰资源丰富，当地新兴蒸馏酒企业正利用这一原料开拓本土烈酒产业，龙舌兰酒被称为印度的「蓝色黄金」，有望成为新的饮料经济增长点。
It is never easy to speak up when a fellow diner says "let's just divide it!"
中文摘要 本文探讨聚餐时朋友提议平摊账单而自己消费较少时如何开口拒绝的社交难题，分析背后的心理与经济考量，提供实用沟通建议。
Investors across Asia have been largely shut out of the world’s largest-ever initial public offering, which has forced them to find creative ways to make bets on SpaceX’s $75 billion global spectacle.
中文摘要 亚洲投资者因市场限制被大规模排除在SpaceX 750亿美元IPO认购之外，被迫寻找另类途径间接布局，包括投资相关基金或二级市场衍生品等方式参与这一史上最大规模IPO。
Brand Industrial Services Inc., a scaffolding and industrial services company that does business as BrandSafway, reported sharply lower first-quarter earnings as higher costs and spending on its push into data center construction weighed on results, according to people familiar with the matter.
中文摘要 脚手架及工业服务公司Brand Industrial Services（旗下品牌BrandSafway）公布第一季度业绩大幅下滑，原因是向数据中心建设领域转型带来的成本上升及相关投入增加，显著压缩了公司利润率。
Rockets-to-AI group priced its shares at $135 in deal that drew blockbuster investor demand
中文摘要 马斯克旗下SpaceX完成全球史上最大规模IPO，融资750亿美元，发行价定为每股135美元，获得投资者踊跃认购。该公司业务涵盖火箭发射及人工智能等领域。
Oil extended declines after President Donald Trump said a peace deal with Iran could be signed as soon as the weekend following US military strikes that cast doubt over progress to end the war.
中文摘要 特朗普表示美伊和平协议最快本周末可望签署，此前美国已对伊朗实施军事打击。受停战预期升温影响，国际油价延续跌势，市场对中东供应中断的风险溢价有所回落。
Asian stocks rose, tracking a rally on Wall Street after President Donald Trump said the US was nearing an agreement with Iran, raising hopes for a diplomatic end to the conflict. Oil dropped.
中文摘要 特朗普暗示美伊即将达成协议，带动华尔街股市上涨，亚洲股市跟涨。与此同时，油价下跌，市场整体风险情绪改善，投资者对中东冲突外交解决的预期明显升温。
5 回复 · 程序员 节点
7 回复 · 程序员 节点
4 回复 · 程序员 节点
10 回复 · 程序员 节点
10 回复 · Apple 节点
45 回复 · Apple 节点
12 回复 · Apple 节点
9 回复 · Apple 节点
28 回复 · Apple 节点
8 回复 · Apple 节点
[!NOTE] Claude Fable 5 & Claude Mythos 5 System Card 如果想深挖，随便找个AI就能挖了 这边又开新坑，那一定是因为我要吐槽！！！！ [!CHECK] 总结 图片总结 文配图总结 [!INFO] 一张图两个字 端妃 嘿呦喂！我就端着，就端着，就端着。 TL;DR 虽然Mythos 5和Fable 5都是新一代超模，性能超越了之前的Mythos Preview，乃们浅浅氪金的屁民们只配用Fable 5，我们还为了人类的福祉，给你们的Fable 5添加了极其先进的外审，传说中的 ASL-3 blocking classifiers，这样在标准用户提
本帖使用社区开源推广，符合推广要求。我申明并遵循社区要求的以下内容： 我的帖子已经打上 开源推广 标签： 是 我的开源项目完整开源，无未开源部分： 是 我的开源项目已链接认可 LINUX DO 社区： 是 我帖子内的项目介绍，AI生成、润色内容部分已截图发出： 是 以上选择我承诺是永久有效的，接受社区和佬友监督： 是 以下为项目介绍正文内容，AI生成、润色内容已使用截图方式发出 剧透 [!warning] 注：这里有一个三分钟使用极简教程，正式使用前推荐看看：【全开源免费！抢先体验属于个人的Easy Research！Obsidian开发者手把手教你三分钟速通NotEMD！-哔哩哔哩】 htt
6月10日，胖东来创始人于东来在企业内部分享会上公开反思经营管理问题。 他表示：“我们公司30年以来，基本上都是使用超越员工期望的薪酬，这样好的一点是能激发大家的热情，不好的一点，产生溺爱。大家其实不值这么多钱，但是给你了这么多钱，让你产生了一个认知上的偏差，偏差就导致自己认为自己就值这么多钱。好的一面是我在公司里面，因为我值这么多钱，我感觉我很骄傲很幸福。不好的一面，就是一旦走出胖东来就完蛋。这可能就是政策上的过度溺爱。” 于东来对胖东来薪酬制度也表达了看法，他表示：“公司里面我期望的结果是，我们薪资层面，无论是人力资源、财务部门或者营运部门，结合起来，我们给公司制定我们相应的比较科学的薪酬
纯自娱自乐，佬友轻喷 赛波 | CYBERBALL 赛波CYBERBALL — 2026世界杯看球伴侣 赛波（CYBERBALL）源于赛博世界波，将实时赛事、AI预测、全球新闻与数据可视化融合为足球时代的数字竞技场。48支参赛队实时赔率、赛程日历、球员档案、小组赛积分榜——一屏尽览。 53 个帖子 - 28 位参与者 阅读完整话题
转自飞书群 注意，内测群貌似不是每个max用户都能进的。 感谢佬友透露内测情况： TengMMVP: 1M上下文窗口，不支持多模态，支持2档思考强度。 44 个帖子 - 38 位参与者 阅读完整话题
今天很多号挂了，我们一直在找号找渠道，所以白天公益站不可用了一段时间 Codex公益站：https://new.sharedchat.cc/ https://rawchat.cn/ 声明：在资源紧张的时候我们还是会优先保障付费站的，希望各位佬友理解 qq群：758607042（有问题群内反馈回复会快点，群内有技术支持） 84 个帖子 - 78 位参与者 阅读完整话题
据传：OpenAI已测试内部代号为kepler和kindle的两个新检查点。kindle-alpha被曝已选为发布候选。 来源公众号 37 个帖子 - 35 位参与者 阅读完整话题
我虽然不高强度冲浪，但是快两年了，也没有到3级。 差浏览量，这个浏览量。我很很难达到啊。 我看到一位3级佬，居然是25年底注册的，已经三级了，这也太肝了吧。 有同样的佬嘛。 126 个帖子 - 111 位参与者 阅读完整话题
如题，第一次在L站过生日，来到L站的这几个月里跟着佬友们学到了很多有用的东西，愿佬友们天天开心，社区越来越好 133 个帖子 - 130 位参与者 阅读完整话题
各位干草铺的老友，请立刻停止YOLO模式，现在我不知道是我自己程序的问题还是奥特曼的问题，已经有佬友给我反馈了GPT回复还像被夺舍的情况，虽然我说过号商再薅我就放毒，但这次真不是我，请大家先放弃YOLO，稍后我会停一会儿公益站。YOLO，也就是bypass那个模式，我现在在外面，来不及写详细，大家能理解就好了，不要完全托管给GPT，能夺舍第一次就能有第二次，至少今天大家先放弃这种模式哈 34 个帖子 - 31 位参与者 阅读完整话题