哔哩哔哩哔哩哔哩新浪新闻新浪新闻新浪网新浪网GitHubGitHubCSDNCSDNIT之家IT之家36氪36氪AcFunAcFunAdafruit BlogAdafruit BlogAIbaseAIbaseAnt Bailing Developer BlogAnt Bailing Developer BlogAnthropicAnthropic小众软件小众软件AppleAppleAppleInsiderAppleInsiderArs TechnicaArs TechnicaBerkeley AI ResearchBerkeley AI ResearchBBC NewsBBC NewsBD Tech TalksBD Tech TalksBerkeley RDIBerkeley RDIBloombergBloombergBoston Dynamics BlogBoston Dynamics BlogByteDance Seed ResearchByteDance Seed ResearchClaude BlogClaude BlogClaude Code ReleasesClaude Code ReleasesCloudflare BlogCloudflare BlogCMU Machine Learning BlogCMU Machine Learning Blog酷安酷安crates.iocrates.ioCrowd SupplyCrowd SupplyCSS-TricksCSS-Tricks51CTO51CTOCult of MacCult of MacCursor BlogCursor BlogDaring FireballDaring FireballDario AmodeiDario AmodeiDeepLearning.AI · The BatchDeepLearning.AI · The BatchGoogle DeepMind BlogGoogle DeepMind BlogDeepSeek GitHubDeepSeek GitHubDescript BlogDescript BlogDEV.toDEV.to数字尾巴数字尾巴Docker HubDocker HubDwarkesh PatelDwarkesh PatelEleutherAI BlogEleutherAI BlogEngadgetEngadgetFlathubFlathubFreebuf · 网络安全Freebuf · 网络安全freeCodeCampfreeCodeCampGary MarcusGary Marcus极客公园极客公园果核剥壳果核剥壳GizmodoGizmodoGoogleGoogleHackadayHackadayHacker NewsHacker NewsHelloGitHubHelloGitHubHomebrewHomebrew全球主机交流全球主机交流Hugging FaceHugging Face虎嗅虎嗅IEEE SpectrumIEEE Spectrum爱范儿爱范儿inclusionAIinclusionAIIndie HackersIndie HackersInfoQInfoQInterconnectsInterconnects稀土掘金稀土掘金Andrej KarpathyAndrej KarpathyKickstarter · TechnologyKickstarter · TechnologyMoonshot AI KimiMoonshot AI KimiLatent SpaceLatent SpaceLil'Log (Lilian Weng)Lil'Log (Lilian Weng)LiliputingLiliputingLinux.doLinux.doLMSYS BlogLMSYS BlogLobstersLobstersLWN.netLWN.netMacRumorsMacRumorsMake: MagazineMake: MagazineMaven CentralMaven CentralMediumMediumMeituan LongCatMeituan LongCatMeta AI BlogMeta AI BlogMeta EngineeringMeta EngineeringMidjourney UpdatesMidjourney UpdatesMiniMaxMiniMaxMIT News · RoboticsMIT News · RoboticsMIT Technology ReviewMIT Technology ReviewMozilla.ai BlogMozilla.ai BlogNew Atlas · RoboticsNew Atlas · Robotics9to5Mac9to5MacNodeSeekNodeSeek牛客牛客NuGetNuGetNVIDIANVIDIAOne Useful ThingOne Useful ThingOpen Robotics BlogOpen Robotics BlogOpenAIOpenAIOpenRouterOpenRouterPackagistPackagist远景论坛远景论坛PhoronixPhoronixPlanet ROSPlanet ROS吾爱破解吾爱破解Product HuntProduct HuntPyPIPyPI量子位 · 具身智能量子位 · 具身智能QoderQoderQwenQwenRedditRedditRFC EditorRFC EditorRobohubRobohubThe Robot ReportThe Robot ReportRobotics & Automation NewsRobotics & Automation NewsRobotics TomorrowRobotics TomorrowROS DiscourseROS Discourse阮一峰的网络日志阮一峰的网络日志RubyGemsRubyGemsRunwayRunwaySam AltmanSam AltmanAhead of AIAhead of AISecurityOnlineSecurityOnlineServeTheHomeServeTheHomeServiceNow AIServiceNow AISimon WillisonSimon WillisonSlashdotSlashdotSmashing MagazineSmashing MagazineSolidotSolidot少数派少数派Stack OverflowStack OverflowStack Overflow BlogStack Overflow BlogSubstackSubstackSuno BlogSuno BlogSynced ReviewSynced Review淘宝逛一逛淘宝逛一逛TechCrunchTechCrunchTech Xplore · RoboticsTech Xplore · RoboticsTencent Hunyuan ResearchTencent Hunyuan ResearchThe DecoderThe DecoderThe GradientThe GradientThe RegisterThe RegisterThe VergeThe VergeTindie BlogTindie BlogTomer TunguzTomer TunguzTom's HardwareTom's HardwareTransformer CircuitsTransformer Circuits优设网优设网UiverseUiverseV2EXV2EXVentureBeat · AIVentureBeat · AIWiredWiredxAI NewsxAI News小鹅通小鹅通Zhipu AI ResearchZhipu AI ResearchCISA KEVCISA KEVNVDNVDOSV.devOSV.devTelegram OSINTTelegram OSINT

科技

198 个榜单
Ars Technica
实时热榜
12分钟前更新
共 20 条热搜查看完整榜单
2小时前更新
  • 01
    From CUDA to MLX: How K-Search Brings Decades of Kernel Expertise to Apple Silicon
    Figure 1: CUDA-to-MLX optimization translation map. CUDA optimization knowledge can be translated into architecture-native MLX strategies rather than copied instruction-for-instruction. We face a new epoch in computing. Hardware is changing rapidly — not just faster GPUs, but a growing range of chips from different vendors, each with its own architecture and often tailored to specific AI workloads. Software is changing just as fast, and AI coding tools now generate in minutes what took months of
  • 02
    Teaching LLMs to Update Beliefs for Efficient Long-Horizon Interaction
    Overview of ABBEL compared to traditional recursive summarization. Beliefs replace the full interaction history as the agent’s working context, and belief grading improves performance by supervising the contents of each belief state.. As task horizons grow, LLM contexts can’t scale forever. Self-summarization enables concise, interpretable contexts, but at a significant performance cost, especially for human assistance domains where high quality data is scarce, e.g., collaborative code generatio
  • 03
    Intelligence is Free, Now What? <br> Data Systems for, of, and by Agents
    ... government of the people, by the people, for the people ... — Abraham Lincoln, Gettysburg Address (1863) The cost of AI is dropping rapidly. GPT-4-class capabilities cost roughly $30 per million tokens in early 2023; today the same runs under $1 , and some providers are pushing costs below $0.10 . Across benchmarks, inference prices have fallen between 9x and 900x per year , with a median decline near 50x. Even frontier models are getting dramatically cheaper each generation, with open-sourc
  • 04
    2026 BAIR Graduate Showcase
    Congratulations to the Berkeley Artificial Intelligence Research (BAIR) Lab class of 2026! This year, BAIR celebrates another remarkable group of Ph.D. graduates whose curiosity, creativity, and perseverance have pushed the frontiers of artificial intelligence and machine learning. Their work spans the breadth of modern AI — robotics and embodied intelligence, large language models and reasoning, computer vision, generative modeling, AI safety, human-AI interaction, AI for science and healthcare
  • 05
    Adaptive Parallel Reasoning: The Next Paradigm in Efficient Inference Scaling
    Overview of adaptive parallel reasoning. What if a reasoning model could decide for itself when to decompose and parallelize independent subtasks, how many concurrent threads to spawn, and how to coordinate them based on the problem at hand? We provide a detailed analysis of recent progress in the field of parallel reasoning, especially Adaptive Parallel Reasoning. Disclosure: this post is part landscape survey, part perspective on adaptive parallel reasoning. One of the authors (Tony Lian) co-l
  • 06
    Gradient-based Planning for World Models at Longer Horizons
    GRASP is a new gradient-based planner for learned dynamics (a “world model”) that makes long-horizon planning practical by (1) lifting the trajectory into virtual states so optimization is parallel across time, (2) adding stochasticity directly to the state iterates for exploration, and (3) reshaping gradients so actions get clean signals while we avoid brittle “state-input” gradients through high-dimensional vision models. Large, learned world models are becoming increasingly capable. They can
  • 07
    Identifying Interactions at Scale for LLMs
    Understanding the behavior of complex machine learning systems, particularly Large Language Models (LLMs), is a critical challenge in modern artificial intelligence. Interpretability research aims to make the decision-making process more transparent to model builders and impacted humans, a step toward safer and more trustworthy AI. To gain a comprehensive understanding, we can analyze these systems through different lenses: feature attribution , which isolates the specific input features driving
  • 08
    Information-Driven Design of Imaging Systems
    An encoder (optical system) maps objects to noiseless images, which noise corrupts into measurements. Our information estimator uses only these noisy measurements and a noise model to quantify how well measurements distinguish objects. Many imaging systems produce measurements that humans never see or cannot interpret directly. Your smartphone processes raw sensor data through algorithms before producing the final photo. MRI scanners collect frequency-space measurements that require reconstructi
  • 09
    RL without TD learning
    In this post, I’ll introduce a reinforcement learning (RL) algorithm based on an “alternative” paradigm: divide and conquer . Unlike traditional methods, this algorithm is not based on temporal difference (TD) learning (which has scalability challenges ), and scales well to long-horizon tasks. We can do Reinforcement Learning (RL) based on divide and conquer, instead of temporal difference (TD) learning. Problem setting: off-policy RL Our problem setting is off-policy RL . Let’s briefly review w
  • 10
    What exactly does word2vec learn?
    What exactly does word2vec learn, and how? Answering this question amounts to understanding representation learning in a minimal yet interesting language modeling task. Despite the fact that word2vec is a well-known precursor to modern language models, for many years, researchers lacked a quantitative and predictive theory describing its learning process. In our new paper , we finally provide such a theory. We prove that there are realistic, practical regimes in which the learning problem reduce
共 10 条热搜查看完整榜单
BD Tech Talks
实时热榜
2小时前更新
共 30 条热搜查看完整榜单
Berkeley RDI
实时热榜
5小时前更新
共 30 条热搜查看完整榜单
BBC News
科技
15分钟前更新
共 21 条热搜查看完整榜单
2小时前更新
  • 01
    3 Takeaways After Debuting Stretch at MODEX 2022 | Boston Dynamics
    Boston Dynamics commercially launched Stretch, our autonomous robot for the warehouse, at MODEX 2022 in Atlanta in late March. At demos in the booth each day, crowds watched as the robots demonstrated one of the most common and difficult to solve challenges in logistics today: truck unloading.mrodin@bostondynamics.com (Boston Dynamics Team)
  • 02
    Doing More with Spot | Boston Dynamics
    In addition to Spot’s revolutionary mobility, one of its most important attributes is that the robot is designed as a scalable platform that is constantly evolving and expanding on the number of dull, dirty and dangerous jobs it can handle for our customers.mrodin@bostondynamics.com (Boston Dynamics Team)
  • 03
    3 Ways Robots See the World | Boston Dynamics
    Although we are all familiar with the five senses humans use to perceive our surroundings, for robots, these senses look a little different. Robots are rapidly becoming more prevalent in our everyday lives which can take a plethora of forms, from industrial applications, to delivery drones to robotic waiters. No matter what the role, these robots require sensors of all sorts to see the environment around them. The combination of cameras, LiDAR and gas sensors, together play a vital role in inforDonovan Chew
  • 04
    How Much Radiation Can Spot Withstand? | Boston Dynamics
    Limiting exposure to radiation is a top priority for nuclear generators, and organizations are looking increasingly to agile mobile robots to keep their workforce safe. Boston Dynamics Spot allows customers to automate inspection tasks and map the presence of radiation on site. Additionally, operators are able to perform manual inspections remotely, either using the tablet controller from a safe distance, or Scout desktop software.mrodin@bostondynamics.com (Boston Dynamics Team)
  • 05
    What Makes an Effective Research Robot? | Boston Dynamics
    Robots are used for a host of different research topics and applications, from studying human-robot interaction to collecting samples in inhospitable environments to developing new commercial applications of robotics and artificial intelligence. All of these different research goals may require unique capabilities or specifications, but there are some consistent factors for what makes an effective robot in a research context. First and foremost, a research robot needs to serve the research; a romrodin@bostondynamics.com (Boston Dynamics Team)
  • 06
    A New Approach to Predictive Maintenance Challenges | Boston Dynamics
    Effective maintenance programs are vital in manufacturing, energy and utilities, and other asset-intensive industries—preventing equipment failures that lead to costly downtime and repairs. Smart organizations have shifted toward predictive maintenance strategies. Instead of periodically checking assets and replacing parts on fixed schedules, companies are now using artificial intelligence (AI) to identify warning signs of impending issues before they cause damage or outages. This approach can hBob Ochiai
  • 07
    What to Do with a Legged Robot in Academia and Research
    Spot’s easy mobility, along with a host of other benefits, has made it a popular choice in academia and research institutions. The ready-out-of-the-box quadruped robot helps researchers focus on their objectives, explore tough problems, and experiment with the many ways robotics can positively impact people’s lives. Here are just a few ways that educational and research organizations are working with Spot today.mrodin@bostondynamics.com (Boston Dynamics Team)
  • 08
    Unlock the Data that Matters Most
    As a quadruped robot, Spot can operate virtually anywhere a person can—going up and down stairs, indoors or outdoors, and through cramped or confined spaces. While this mobility can be impressive on its own, its real value is in automating sensing and inspection in areas inaccessible with wheeled or tracked robots and drones. You can send Spot into areas designed for people, that are unstructured, or even hazardous, to collect the data you need to understand what's going on in your facility or amrodin@bostondynamics.com (Boston Dynamics Team)
  • 09
    Break Through Supply Chain Blocks with Automated Container Unloading
    Customers finding bare spots on grocery store shelves or waiting months instead of weeks for a furniture delivery are feeling the ripple effects of supply chain disruptions—with more shortages and delays anticipated as the holiday season approaches. Much of the merchandise is out there, stalled in cargo ports loaded with shipping containers at unprecedented volumes. Labor shortages have slowed the pace of unloading containers and warehouses to a crawl, and ports on both U.S. coasts have requiredmrodin@bostondynamics.com (Boston Dynamics Team)
  • 10
    What is Dynamic Sensing?
    Artificial intelligence is changing the way businesses operate. AI systems demand a continuous flow of repeatable data that’s difficult to collect on physical sites. Today, the health of equipment in facilities is monitored through a mix of fixed sensors and operator rounds. These sensors are limited in perspective, expensive to scale, and can’t adapt to changes. Companies make up the data deficit by spending human resources on tedious data collection. But people are too inconsistent at data colmrodin@bostondynamics.com (Boston Dynamics Team)
共 10 条热搜查看完整榜单
Bloomberg
科技
1分钟前更新
  • 01
    Counterpoint's Shah on Nvidia, Circular Deals
    Neil Shah, Vice President at Counterpoint, discusses the outlook for Nvidia and the circular deals underpinning the AI boom. He speaks with Shery Ahn on Bloomberg: The Asia Trade. (Source: Bloomberg)
  • 02
    AAPA on Asia's Airline Industry, Passenger Demand
    Wong Hong, Director General at the Association of Asia Pacific Airlines, discusses the outlook for the region's carriers and the impact from higher jet fuel prices on passenger demand. He speaks with Haidi Stroud-Watts on Bloomberg: The Asia Trade. (Source: Bloomberg)
  • 03
    South Korea's AI Investment Plans
    South Korea expects to deploy more than 1 trillion won in fresh capital next year into a new sovereign wealth fund targeting AI and other strategic industries, Min Kyung-seol, the country's deputy finance minister for innovation and growth, tells Bloomberg Television's Shery Ahn in an exclusive interview. (Source: Bloomberg)
  • 04
    Peak Memory Fears Put Focus on Samsung, SK Hynix Cash Returns
    Memory-chip makers may have found a new defense against the inevitable cooling of their boom: giving more cash back to shareholders.Sunny Bangia and Abhishek Vishnoi
  • 05
    Korea Sovereign Wealth Fund to Join Global Race for AI, Robotics
    South Korea expects to deploy more than 1 trillion won ($707 million) of fresh capital into a new sovereign wealth fund targeting AI and other strategic industries next year, joining a global push by governments to mobilize investment and gain an edge in high-tech sectors.Jaehyun Eom and Shery Ahn
  • 06
    Redpoint’s Brescia on Nvidia’s $500B Funding, AI Capex
    Erica Brescia, Managing Director at Redpoint Ventures, discussed Super Micro's robust business model and strong market position during the segment. Despite recent volatility, Super Micro benefits from significant pricing power and a durable business model as a key supplier of hardware components essential for AI-driven growth. She speaks with Romaine Bostick on "The Close." (Source: Bloomberg)
  • 07
    Upwork CEO on AI Disruption Concerns, Earnings Slide
    Hayden Brown, CEO of Upwork, discusses the company's recent financial performance after its full-year forecast missed analyst estimates, leading to a significant drop in the stock price. Brown discusses the sources of pressure facing Upwork and how the company plans to adapt to these challenges. She speaks with Romaine Bostick on "The Close." (Source: Bloomberg)
  • 08
    Accel Raises $3.5 Billion to Invest in Emerging Global AI Startups
    Accel, a global venture capital firm with offices in Silicon Valley, London and Bangalore, India, has raised $3.5 billion in new funds for early-stage investments across the world.Natasha Mascarenhas
  • 09
    Google to Add Three Subsea Cable Systems Across the Americas
    Alphabet Inc.’s Google announced new cable links in the Americas on Tuesday, including one between Panama, the Dominican Republic and Chile, which is trying to establish itself as a regional digital hub.Antonia Mufarech
  • 10
    CoreWeave Took a Step in the Right Direction, Gil Luria Says
    Gil Luria, head of technology research at D.A. Davidson, reacts to CoreWeave topping analysts' estimates for second-quarter sales after signing up more customers in a booming market, with sales more than doubling to $2.58 billion. He speaks on "Bloomberg The Close." (Source: Bloomberg)
  • 11
    SiTime CEO on Role in AI Data Centers, Renesas Deal
    Rajesh Vashist, Chairman and CEO of SiTime, discussed the company's impressive triple-digit year-over-year revenue growth, attributing much of the acceleration to AI-driven demand and the expanding data center market. SiTime's technology plays a critical role in timing synchronization across various data center components, including TPUs, GPUs, switches, accelerator cards, optical modules, and cables, ensuring high performance and cost efficiency. He speaks with Romaine Bostick & Isabelle Lee on
  • 12
    Vertiv CEO on AI Infrastructure Buildout, Supply Chain
    Gio Albertazzi, CEO of Vertiv, joined the program to discuss the company's recent earnings report. Unlike many AI-related companies that often beat revenue expectations but miss on profitability or margins, Vertiv reported the opposite: it beat on margins and profitability but missed on revenue. He speaks with Romaine Bostick & Isabelle Lee on "The Close." (Source: Bloomberg)
  • 13
    Moonshot AI’s Kimi K3: A New Contender in the US-China AI Race: Podcast
    Chinese startup Moonshot has put the AI world on notice. On today’s Big Take Asia podcast: What Kimi K3 means for the future of the global race for AI supremacy.K Oanh Ha and Yang Yang
  • 14
    AI Infrastructure Boom Drives Market Movers | Closing Bell
    Comprehensive cross-platform coverage of the U.S. market close on Bloomberg Television, Bloomberg Radio, and YouTube with Romaine Bostick, Isabelle Lee, Kristine Aquino and Tim Stenovec. (Source: Bloomberg)
  • 15
    CoreWeave Shares Surge After Booming AI Demand Bolsters Outlook
    CoreWeave Inc., a provider of computing that powers artificial intelligence systems, soared in late trading after the AI spending frenzy spurred faster sales growth than anticipated.Dina Bass and Brody Ford
  • 16
    Super Micro Gives Sales Outlook That Tops Rosiest Projection
    Super Micro Computer Inc. gave a revenue forecast for the current quarter that topped analysts’ estimates, a sign the booming artificial intelligence market continues to bolster sales of the company’s servers.Dina Bass
  • 17
    Bitcoin Wallet Hack Shakes Industry | Bloomberg Crypto 8/11/2026
    "Bloomberg Crypto" covers the people, transactions, and technology shaping the world of decentralized finance. Today's guests: Galaxy Digital Head of Firmwide Research Alex Thorn, ledger Chief Human Agency officer Ian Rogers, Bitwise Asset Management CIO Matt Hougan, Robinhood Crypto General Manager Johann Kerbrat, and AnchorWatch COO & Co-Founder Becca Amilee Rubenfeld. (Source: Bloomberg)
  • 18
    Bitcoin Will 'Survive' Cold Wallet Hack, Says Galaxy Digital's Thorn
    Alex Thorn, head of firmwide research at Galaxy Digital, joins Isabelle Lee and Tim Stenovec on "Bloomberg Crypto." The crypto community is reeling from a breach involving "cold" Bitcoin wallets made by Coinkite. Hackers were able to steal an estimated $130 million worth of Bitcoin from about 7,300 wallet addresses. (Source: Bloomberg)
  • 19
    Sequoia Bets on Neros for American Drone Future (Correct)
    Drone startup Neros has raised $250 million in a round co-led by Sequoia that triples its valuation to $2.5 billion as it expands beyond strike drones into autonomous systems and drone interceptors. CEO Soren Monroe-Anderson and Sequoia partner Shaun Maguire join Bloomberg to discuss plans to ramp up drone production, the race to build a US supply chain independent of China, and why they ultimately see a path to producing one million drones a year. They join Ed Ludlow on "Bloomberg Tech." (Corre
  • 20
    Google Team Tells Applicants Its HR Filters Are Unreliable
    Alphabet Inc.’s Google pitches its artificial intelligence tools to corporate clients as a way to more quickly sift through a mountain of job applications to find the most promising candidates. Some of its own AI researchers don’t want to rely on these tools when recruiting.Julia Love
共 20 条热搜查看完整榜单
15分钟前更新
共 30 条热搜查看完整榜单
Claude Blog
实时热榜
12分钟前更新
共 30 条热搜查看完整榜单
45分钟前更新
  • 01
    v2.1.228
    What's changed Fixed interactive sessions that could stop redrawing entirely, while the process kept running, after a rare internal layout error Fixed git / Git Bash not being found on Windows when Claude Code is launched from a parent folder of the git installation Fixed /tui reverting the session to an earlier model when /model had been changed since the last response Fixed cross-session messaging sometimes starting without an inbox in the first session after install or upgrade Fixed Remote Coashwin-ant
  • 02
    v2.1.227
    What's changed Fixed feature flags being evaluated without the user's subscription tier when a session started with an expired login token, which could wrongly prompt Max plan users to enable usage credits for Fable Fixed every Bash command failing under claude-code-action with allowed_non_write_users on GitHub-hosted runners Fixed /tui bringing back a conversation that had been rewound to before its first message Improved slash-command menu: blue now marks only the selected row, matched charactashwin-ant
  • 03
    v2.1.226
    What's changed Bug fixes and reliability improvementsashwin-ant
  • 04
    v2.1.225
    What's changed Added gateway spend-limit support to Claude Code's usage warning; the limit-reached message now names the cap, its reset time, and the operator's message (requires the gateway on 2.1.225) Added a workspace trust prompt to claude agents for untrusted directories, matching the behavior of claude Fixed a transient 401 replacing a long-lived CLAUDE_CODE_OAUTH_TOKEN with a stored login's short-lived token, breaking headless sessions until restart Fixed MCP OAuth servers on macOS intermashwin-ant
  • 05
    v2.1.224
    What's changed Added self-hosted environments: claude self-hosted-runner turns your own machines or containers into a place Claude Code web, mobile, and desktop sessions can run, on Team and Enterprise plans Added archive plugin source: install plugins from a zip over HTTPS without git or npm, with optional SHA-256 pinning Added a cancel-and-confirm step when removing an unavailable paste changes a command's text Added ANTHROPIC_BEDROCK_REGION_PREFIX env var for Bedrock to prefer a specific crosashwin-ant
  • 06
    v2.1.223
    What's changed Added owner wildcard entries ( "owner/*" ) to the strictKnownMarketplaces and blockedMarketplaces managed settings for allowing or blocking all marketplace repos under a GitHub org Added a warning when workflow agents, forked skills, slash commands, or resumed background agents' requested subagent model is restricted and the parent model runs instead Added a /teleport hint in cloud sessions showing how to continue locally with claude --teleport <session id> Fixed a Bash permissionashwin-ant
  • 07
    v2.1.222
    What's changed Fixed worktree-isolated sessions and their subagents being able to run destructive git commands against the main checkout; isolation now applies to file edits and Bash in every session type Fixed PreToolUse auto-allow hooks bypassing tool restrictions in background agent tasks (summaries, compaction, renames) Fixed /usage-credits on Team and Enterprise showing "you've already sent a usage credit request" for members whose earlier request was dismissed, blocking them from sending aashwin-ant
  • 08
    v2.1.221
    What's changed [VSCode] Added Focus view: a chat-menu toggle that hides tool activity behind an expandable per-turn summary with a live running-tool indicator, toggled with Ctrl+Alt+F or the "Claude Code: Toggle Focus view" command Added mode: "mask" for sandbox credential files on Linux and WSL — sandboxed commands read a sentinel copy (the whole file, or just the spans captured by an extract regex) while the sandbox proxy substitutes the real value on egress; on macOS file masking falls back tashwin-ant
  • 09
    v2.1.220
    What's changed Bug fixes and reliability improvementsashwin-ant
  • 10
    v2.1.219
    What's changed Added Claude Opus 5 ( claude-opus-5 ), now the default Opus model — 1M context, fast mode at $10/$50 per Mtok Added sandbox.network.strictAllowlist setting to deny non-allowlisted hosts for sandboxed commands without prompting Added DirectoryAdded hook that fires after /add-dir or the SDK register_repo_root control request registers a new working directory mid-session Added mcp_server_errors to the headless stream-json init event, listing --mcp-config entries skipped by config valashwin-ant
共 10 条热搜查看完整榜单
Cloudflare Blog
实时热榜
41分钟前更新
共 20 条热搜查看完整榜单
2小时前更新
  • 01
    Forking-Sequences — Part II: Multi-Horizon Forecast Ensembling with Reduced Volatility
    Based on: Potosnak, W., Wolff, M., Cao, M., Ma, R., Konstantinova, T., Efimov, D., Mahoney, M.W., Oreshkin, B., & Olivares, K.G. "Forking-Sequences: Statistically and Computationally Efficient Multi-Horizon Forecasting with Reduced Volatility." Transactions on Machine Learning Research, 2026. (Disclaimer: Code implementation not used in the paper; not affiliated with Amazon — provided as a reference for forking-sequences and forecast ensembling) TL;DR Ensembling, nearly for free. Forking-sequencWilla Potosnak
共 1 条热搜查看完整榜单