微博微博今日头条今日头条百度百度抖音抖音快手快手哔哩哔哩哔哩哔哩知乎知乎腾讯新闻腾讯新闻网易新闻网易新闻澎湃新闻澎湃新闻新浪新闻新浪新闻新浪网新浪网豆瓣电影豆瓣电影GitHubGitHubCSDNCSDNIT之家IT之家36氪36氪宝庆银楼宝庆银楼中国黄金中国黄金周生生周生生英皇珠宝英皇珠宝老凤祥（广东）老凤祥（广东）六福珠宝六福珠宝周大福周大福周六福周六福AcFunAcFunAdafruit BlogAdafruit BlogAIbaseAIbaseAnt Bailing Developer BlogAnt Bailing Developer BlogAnthropicAnthropicAP News · TechnologyAP News · Technology小众软件小众软件AppleAppleApple PodcastsApple PodcastsAppleInsiderAppleInsiderArs TechnicaArs TechnicaarXivarXivAxiosAxiosBerkeley AI ResearchBerkeley AI ResearchBarchartBarchartBBC NewsBBC NewsBD Tech TalksBD Tech TalksBerkeley RDIBerkeley RDIBig ThinkBig Think哔哩哔哩热搜哔哩哔哩热搜哔哩哔哩热门视频哔哩哔哩热门视频BinanceBinanceBloombergBloombergBlueskyBlueskyBoston Dynamics BlogBoston Dynamics BlogBusiness InsiderBusiness InsiderByteDance Seed ResearchByteDance Seed Research参考消息参考消息虫部落虫部落Claude BlogClaude BlogClaude Code ReleasesClaude Code ReleasesCloudflare BlogCloudflare BlogCMU Machine Learning BlogCMU Machine Learning BlogCoinDeskCoinDeskCoinGeckoCoinGeckoCointelegraphCointelegraph酷安酷安crates.iocrates.ioCrowd SupplyCrowd SupplyCSS-TricksCSS-Tricks51CTO51CTOCult of MacCult of MacCursor BlogCursor Blogdaily.dev · Populardaily.dev · PopularDaring FireballDaring FireballDario AmodeiDario AmodeiDeepLearning.AI · The BatchDeepLearning.AI · The BatchGoogle DeepMind BlogGoogle DeepMind BlogDeepSeek GitHubDeepSeek GitHubDefiLlamaDefiLlamaDescript BlogDescript BlogDEV.toDEV.to数字尾巴数字尾巴Docker HubDocker Hub豆瓣读书豆瓣读书豆瓣讨论豆瓣讨论Dwarkesh PatelDwarkesh Patel中国地震台中国地震台The EconomistThe EconomistEleutherAI BlogEleutherAI BlogEngadgetEngadgetFinancial TimesFinancial TimesFlathubFlathubFreebuf · 网络安全Freebuf · 网络安全freeCodeCampfreeCodeCampFrontiersFrontiersGameRes 游资网GameRes 游资网Gary MarcusGary Marcus极客公园极客公园原神原神果核剥壳果核剥壳GizmodoGizmodoGoogleGoogleGoogle TrendsGoogle Trends国家法律法规数据库国家法律法规数据库中国政府网中国政府网果壳果壳HackadayHackadayHacker NewsHacker NewsHelloGitHubHelloGitHub历史上的今天历史上的今天HomebrewHomebrew崩坏3崩坏3全球主机交流全球主机交流Hugging FaceHugging Face活动行活动行虎扑虎扑虎嗅虎嗅IEEE SpectrumIEEE Spectrum爱范儿爱范儿凤凰网 · 热点资讯凤凰网 · 热点资讯inclusionAIinclusionAIIndie HackersIndie HackersInfoQInfoQInterconnectsInterconnects爱奇艺热播榜爱奇艺热播榜IT之家「喜加一」IT之家「喜加一」简书简书稀土掘金稀土掘金靠谱新闻靠谱新闻Andrej KarpathyAndrej KarpathyKickstarter · TechnologyKickstarter · TechnologyMoonshot AI KimiMoonshot AI KimiLatent SpaceLatent SpaceLessWrongLessWrongLichessLichessLil'Log (Lilian Weng)Lil'Log (Lilian Weng)LiliputingLiliputingLinux.doLinux.doLMSYS BlogLMSYS BlogLobstersLobsters英雄联盟英雄联盟LWN.netLWN.netMacRumorsMacRumorsMake: MagazineMake: MagazineMaven CentralMaven CentralMediumMediumMeituan LongCatMeituan LongCatMeta AI BlogMeta AI BlogMeta EngineeringMeta EngineeringMidjourney UpdatesMidjourney Updates工业和信息化部工业和信息化部MiniMaxMiniMaxMIT News · RoboticsMIT News · RoboticsMIT Technology ReviewMIT Technology Review米游社米游社Mozilla.ai BlogMozilla.ai BlogNatureNatureNeuroscience NewsNeuroscience NewsNew Atlas · RoboticsNew Atlas · RoboticsThe New YorkerThe New Yorker水木社区水木社区NGANGA9to5Mac9to5MacNodeSeekNodeSeek牛客牛客NuGetNuGetNVIDIANVIDIA纽约时报纽约时报OEISOEISOne Useful ThingOne Useful ThingOpen Robotics BlogOpen Robotics BlogOpenAIOpenAIOpenAlexOpenAlexOpenReviewOpenReviewOpenRouterOpenRouterPackagistPackagist远景论坛远景论坛PhoronixPhoronixPhys.orgPhys.orgPixivPixivPlanet ROSPlanet ROS吾爱破解吾爱破解PoliticoPolitico中国电建阳光采购网中国电建阳光采购网Product HuntProduct HuntPubMedPubMedPyPIPyPI量子位 · 具身智能量子位 · 具身智能QoderQoder腾讯视频热搜榜腾讯视频热搜榜Quanta MagazineQuanta MagazineQwenQwenRedditRedditReutersReutersRFC 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 AltmanScienceAlertScienceAlertAhead of AIAhead of AISecurityOnlineSecurityOnlineServeTheHomeServeTheHomeServiceNow AIServiceNow AISimon WillisonSimon WillisonSingularity HubSingularity HubSky NewsSky NewsSlashdotSlashdotSmashing MagazineSmashing Magazine什么值得买什么值得买SolidotSolidotSpotifySpotify俄罗斯卫星通讯社俄罗斯卫星通讯社少数派少数派Stack OverflowStack OverflowStack Overflow BlogStack Overflow Blog崩坏：星穹铁道崩坏：星穹铁道SteamSteamSubstackSubstackSunoSunoSuno BlogSuno BlogSynced ReviewSynced Review淘宝逛一逛淘宝逛一逛TechCrunchTechCrunchTech Xplore · RoboticsTech Xplore · Robotics腾讯热点腾讯热点Tencent Hunyuan ResearchTencent Hunyuan ResearchThe AtlanticThe AtlanticThe DecoderThe DecoderThe GradientThe GradientThe GuardianThe GuardianThe RegisterThe RegisterThe VergeThe Verge百度贴吧百度贴吧TikTokTikTokTindie BlogTindie BlogTomer TunguzTomer TunguzTom's HardwareTom's HardwareTransformer CircuitsTransformer CircuitsTVmazeTVmaze优设网优设网UiverseUiverseV2EXV2EXVentureBeat · AIVentureBeat · AI万方数据万方数据中央气象台中央气象台微信读书微信读书WikidataWikidataWikipediaWikipediaWiredWiredThe Wall Street JournalThe Wall Street JournalxAI NewsxAI News小鹅通小鹅通雪球 · 热门股票雪球 · 热门股票YahooYahooYahoo FinanceYahoo FinanceYollomiYollomi有道精品课有道精品课YouTubeYouTube游研社游研社联合早报联合早报知乎日报知乎日报Zhipu AI ResearchZhipu AI Research财新数据通财新数据通央视新闻联播央视新闻联播CelesTrakCelesTrakCISA KEVCISA KEV财联社财联社东方财富东方财富法布财经法布财经富途牛牛富途牛牛格隆汇格隆汇HDXHDX和讯网和讯网Investing.comInvesting.com界面新闻界面新闻金十数据金十数据21财经21财经金融界金融界Launch Library 2Launch Library 2MKTNews · 快讯MKTNews · 快讯NASA EONETNASA EONETNOAA/NWSNOAA/NWSNVDNVDopenFDAopenFDAOSV.devOSV.dev新浪财经新浪财经Spaceflight NewsSpaceflight NewsTelegram OSINTTelegram OSINT同花顺同花顺USAspendingUSAspendingUSGS EarthquakesUSGS Earthquakes华尔街见闻华尔街见闻WHO 疫情通报WHO 疫情通报选股宝选股宝第一财经第一财经

实时热搜

710 个榜单
牛客
热门帖子
7小时前更新
共 20 条热搜查看完整榜单
NuGet
累计下载
10小时前更新
  • 0189.7亿
    Newtonsoft.Json
    排名：1 · 版本：13.0.5-beta1 · Json.NET is a popular high-performance JSON framework for .NET · 累计下载：8,969,352,212 · 标签：jsondotnetfoundation, jamesnk, newtonsoft
  • 0276.5亿
    Microsoft.Extensions.DependencyInjection
    排名：2 · 版本：11.0.0-preview.7.26381.103 · Default implementation of dependency injection for Microsoft.Extensions.DependencyInjection. · 累计下载：7,647,894,598aspnet, dotnetframework, Microsoft
  • 0374.8亿
    Microsoft.Extensions.Logging
    排名：3 · 版本：11.0.0-preview.7.26381.103 · Logging infrastructure default implementation for Microsoft.Extensions.Logging. · 累计下载：7,480,062,737aspnet, dotnetframework, Microsoft
  • 0462.2亿
    System.Text.Json
    排名：4 · 版本：11.0.0-preview.7.26381.103 · Provides high-performance and low-allocating types that serialize objects to JavaScript Object Notation (JSON) text and deserialize JSON text to objects, with UTF-8 support built-in. Also provides types to read... · 累计下载：6,215,738,030dotnetframework, Microsoft
  • 0545.2亿
    Microsoft.Bcl.AsyncInterfaces
    排名：5 · 版本：11.0.0-preview.7.26381.103 · Provides the IAsyncEnumerable<T> and IAsyncDisposable interfaces and helper types for .NET Standard 2.0. This package is not required starting with .NET Standard 2.1 and .NET Core 3.0. · 累计下载：4,520,321,818dotnetframework, Microsoft
  • 0641.2亿
    Azure.Core
    排名：6 · 版本：1.61.0 · This is the implementation of the Azure Client Pipeline · 累计下载：4,120,859,569 · 标签：Microsoft, Azure, Client, Pipelineazure-sdk, Microsoft
  • 0731.2亿
    Serilog
    排名：7 · 版本：4.4.1-dev-02443 · Simple .NET logging with fully-structured events · 累计下载：3,124,867,763 · 标签：serilog, logging, semantic, structuredserilog
  • 0829.6亿
    Microsoft.IdentityModel.Abstractions
    排名：8 · 版本：8.22.0 · A package containing thin abstractions for Microsoft.IdentityModel. · 累计下载：2,956,611,572 · 标签：.NET, Windows, Authentication, Identity, AbstractionsAzureAD, Microsoft
  • 0927.9亿
    System.Drawing.Common
    排名：9 · 版本：11.0.0-preview.7.26381.103 · Provides access to GDI+ graphics functionality. Commonly Used Types: System.Drawing.Bitmap System.Drawing.BitmapData System.Drawing.Brush System.Drawing.Font... · 累计下载：2,790,012,488dotnetframework, Microsoft
  • 1027.0亿
    Microsoft.Identity.Client
    排名：10 · 版本：4.87.1-preview.2 · This package contains the binaries of the Microsoft Authentication Library for .NET (MSAL.NET). MSAL.NET makes it easy to obtain tokens from the Microsoft identity platform for developers (formerly Azure... · 累计下载：2,700,963,692 · 标签：Microsoft, Authentication, Library, MSA, MSAL, B2C, Azure, Active, Directory, AAD, More tagsAzureAD, Microsoft
  • 1125.6亿
    Microsoft.Win32.SystemEvents
    排名：11 · 版本：11.0.0-preview.7.26381.103 · Provides access to Windows system event notifications. Commonly Used Types: Microsoft.Win32.SystemEvents · 累计下载：2,564,816,831dotnetframework, Microsoft
  • 1225.3亿
    System.Diagnostics.EventLog
    排名：12 · 版本：11.0.0-preview.7.26381.103 · Provides the System.Diagnostics.EventLog class, which allows the applications to use the Windows event log service. Commonly Used Types: System.Diagnostics.EventLog · 累计下载：2,526,240,059dotnetframework, Microsoft
  • 1324.8亿
    Microsoft.Extensions.Http
    排名：13 · 版本：11.0.0-preview.7.26381.103 · The HttpClient factory is a pattern for configuring and retrieving named HttpClients in a composable way. The HttpClient factory provides extensibility to plug in DelegatingHandlers that address cross-cutting... · 累计下载：2,483,042,165aspnet, dotnetframework, Microsoft
  • 1424.2亿
    Microsoft.EntityFrameworkCore
    排名：14 · 版本：11.0.0-preview.7.26381.103 · Entity Framework Core is a modern object-database mapper for .NET. It supports LINQ queries, change tracking, updates, and schema migrations. EF Core works with SQL Server, Azure SQL Database, SQLite, Azure... · 累计下载：2,417,980,035 · 标签：Entity, Framework, Core, entity-framework-core, EF, Data, O/RM, EntityFramework, EntityFrameworkCore, EFCoreaspnet, dotnetframework, EntityFramework, Microsoft
  • 1522.4亿
    System.Security.Cryptography.Pkcs
    排名：15 · 版本：11.0.0-preview.7.26381.103 · Provides support for PKCS and CMS algorithms. Commonly Used Types: System.Security.Cryptography.Pkcs.EnvelopedCms · 累计下载：2,239,508,234dotnetframework, Microsoft
  • 1621.7亿
    System.Windows.Extensions
    排名：16 · 版本：11.0.0-preview.7.26381.103 · Provides miscellaneous Windows-specific types Commonly Used Types: System.Security.Cryptography.X509Certificates.X509Certificate2UI System.Security.Cryptography.X509Certificates.X509SelectionFlag · 累计下载：2,172,915,352dotnetframework, Microsoft
  • 1719.7亿
    AWSSDK.Core
    排名：17 · 版本：4.0.101.1 · The Amazon Web Services SDK for .NET - Core Runtime · 累计下载：1,969,744,465 · 标签：AWS, Amazon, cloud, aws-sdk-v4awsdotnet
  • 1819.1亿
    Azure.Identity
    排名：18 · 版本：1.21.0 · Provides APIs for authenticating to Microsoft Entra ID · 累计下载：1,907,015,545 · 标签：Microsoft, Azure, Identity, windowsazureofficial, azureofficialazure-sdk, Microsoft
  • 1918.6亿
    System.Memory.Data
    排名：19 · 版本：11.0.0-preview.7.26381.103 · A lightweight abstraction for a payload of bytes. Provides methods for converting between strings, streams, JSON, and bytes. Commonly Used Types: System.BinaryData · 累计下载：1,863,933,387azure-sdk, dotnetframework, Microsoft
  • 2017.1亿
    Microsoft.Extensions.Diagnostics.Abstractions
    排名：20 · 版本：11.0.0-preview.7.26381.103 · Diagnostic abstractions for Microsoft.Extensions.Diagnostics. Commonly Used... · 累计下载：1,712,347,136dotnetframework, Microsoft
共 20 条热搜查看完整榜单
NVIDIA
Blog
15天前更新
  • 01
    NVIDIA Joins NSF State and Regional AI Hubs Program to Expand AI Research and Education Across the US
    NVIDIA is participating in the U.S. National Science Foundation’s (NSF) State and Regional Artificial Intelligence Infrastructure Hubs program, an effort launching today to expand access to the advanced computing, data, software and expertise needed for AI-enabled research and education. Consistent with the aims of the Genesis Mission, the program will support state and multistate groups […]John Josephakis
  • 02
    NVIDIA Alpamayo 2 Super, the Frontier Open Model for Robotaxis and Autonomous Vehicles, Now Available for Commercial Use
    For robotaxis and other autonomous vehicles (AVs), the hardest problems aren’t the everyday scenarios. They’re the rare, complex situations that are difficult to anticipate and train for. Handling these long‑tail events takes more than just object detection and motion prediction. AVs must understand the situation, reason about cause and effect, choose the right action and […]Jessica Soares
  • 03
    As AI Increases Demands on Memory, Storage Steps Up
    Surging AI demands are driving the need for massive datasets and context windows that burst past the confines of system memory. But rising needs aren’t met by simply adding more storage capacity. What’s needed is useful, grounded insights from AI factories and efficient, secure storage architectures that enable those insights. At this week’s Future of […]Jason Hardy
  • 04
    AI Leaders Propose SAFE Guidelines for Cybersecurity Transparency
    Members of the Open Secure AI Alliance — now more than 120 organizations strong — are developing new guidelines to strengthen agentic AI cybersecurity as the annual Black Hat conference begins in Las Vegas today. The Linux Foundation today shared a Request for Comments on Shared AI Findings Exchange (SAFE), a proposed set of guidelines […]Justin Boitano
  • 05
    Best in Class: Stream PC Games and Study on the Same Laptop With GeForce NOW
    Back to school means balancing assignments, deadlines and downtime. GeForce NOW makes it easy to have it all. With cloud gaming, everyday laptops used for class can also become GeForce RTX-powered gaming setups. When it’s time to switch from studying to gaming, members can jump into Halo: Campaign Evolved, as GeForce NOW is bringing one […]GeForce NOW Community
  • 06
    Powerful Compute So Compact, It’s Clutch — Build AI Anywhere With NVIDIA Jetson
    As a discerning AI investor who values style and substance, Sarah Guo knows this season’s standout accessory isn’t the latest designer purse — but what’s inside it. In a recent video, Guo, founder of AI-native venture capital firm Conviction and co-host of the AI podcast No Priors, highlighted how the NVIDIA Jetson platform for edge […]Matthew Leib
  • 07
    Industry Leaders Unite in Open Secure AI Alliance for AI Safety and Security
    Open source software is a critical pillar of the global economy. It underpins cloud computing, financial services, manufacturing, telecommunications, government and internet services by making technology accessible and observable to communities of experts. Cybersecurity is among the top three beneficiaries of open source software. The Open Secure AI Alliance — building on the leadership of […]NVIDIA
  • 08
    NVIDIA Harnesses Vera CPU to Speed Up Design of Next-Generation CPUs and GPUs
    The complexity of modern chip design continues to grow as engineering teams work to develop increasingly sophisticated CPUs, GPUs and AI systems. To help meet that challenge, NVIDIA is collaborating with industry leaders Cadence and Synopsys to optimize critical electronic design automation (EDA) applications for the NVIDIA Vera CPU. NVIDIA is now deploying Vera across […]Ivan Goldwasser
  • 09
    At AI Summit, South Korea Outlines Its AI Future With NVIDIA and Partners
    At this week’s AI Summit in San Francisco, South Korean President Jae Myung Lee and some of the country’s top business leaders and researchers are meeting with NVIDIA and ecosystem partners to chart Korea’s AI progress. Building on NVIDIA founder and CEO Jensen Huang’s visit to Korea last month, this week’s discussions and announcements advance […]NVIDIA Writers
  • 10
    GeForce NOW Sets Sail With ‘Path of Exile: Curse of the Allflame’ Joining the Cloud
    Lock in and load up the cloud. GFN Thursday brings fresh updates and new adventures, all ready to play without waiting for downloads. Set sail in Path of Exile: Curse of the Allflame and charge in Battlefield 6 Season 4 both launching major content for members this week. Then revisit Capcom legends like Breath of […]GeForce NOW Community
  • 11
    NVIDIA AI Supercomputer Comes Online at Naval Postgraduate School
    NVIDIA founder and CEO Jensen Huang today visited the Naval Postgraduate School in Monterey, California, to commission an NVIDIA DGX GB300 system — bringing one of the world’s most powerful AI platforms fully online for the students, researchers and faculty at the U.S. military’s flagship graduate university. “Our nation depends on our men and women […]NVIDIA Writers
  • 12
    NVIDIA Open Sources First GPU-Accelerated Medical Physics Simulation Framework
    Before a healthcare robot can be useful in the real world, it has to learn how the physical world pushes back. Anatomy varies. Instruments bend, press, slip and interact with tissue. Imaging can be noisy or incomplete. And the rare, edge scenarios developers most need to understand don’t appear on schedule. That creates one of […]David Niewolny
  • 13
    Built in Fort Worth: Wistron Opens Advanced Manufacturing Plant to Produce NVIDIA AI Systems
    The AI era runs on AI infrastructure. Many of these advanced systems are built and tested in Texas. Wistron opened its first U.S. manufacturing facility today in Fort Worth — a 324,000-square-foot greenfield plant producing superchips at the heart of some of the world’s most capable AI systems. In front of an audience of Wistron […]NVIDIA Writers
  • 14
    NVIDIA Vera Rubin Driving Performance Per Watt, Lowest Token Cost for Partners Worldwide
    NVIDIA Vera Rubin is here, and it’s going gigascale. Vera Rubin NVL72 production is ramping up with racks running at partners CoreWeave, Google Cloud, Microsoft Azure, Oracle Cloud Infrastructure and Nebius. Spanning 350+ factory sites in 30 countries, Vera Rubin has the largest, most mature rack-scale supply chain ever assembled to meet customer compute demand. […]NVIDIA Writers
  • 15
    Built for Vera Rubin, NVIDIA Spectrum-6 Arrives in Gigascale AI Factories
    AI has entered the gigascale era. The world’s most advanced AI factories are bringing together hundreds of thousands of GPUs and CPUs to train frontier models, power agentic AI and generate intelligence at unprecedented scale. At this level, networking becomes a critical computing power multiplier in driving token generation. Marking a networking milestone, NVIDIA Spectrum-6 […]Scot Schultz
  • 16
    NVIDIA and Partners Build in America, for America
    NVIDIA and its partners are investing in American manufacturing, supply chains, energy grids and skilled workforces so the U.S. can produce the infrastructure needed for better healthcare, breakthrough scientific discovery, stronger industrial productivity and global technology leadership.NVIDIA
  • 17
    At SIGGRAPH, NVIDIA Advances Graphics and Simulation With Agentic and Physical AI
    From open models to real-time simulation, AI and graphics breakthroughs are transforming media, content creation and robotics.NVIDIA Writers
  • 18
    Bristol Myers Squibb Building Life Science Industry’s Most Advanced AI Factory on NVIDIA Vera Rubin
    Erin Davis calls it the “SuperDuperPOD.” That’s two things in one name: pharmaceutical giant Bristol Myers Squibb (BMS) already runs one of the largest AI clusters in life sciences, with serious results to show for it. And they’re doubling down. BMS announced today it is deploying its second NVIDIA DGX SuperPOD, this one built on […]Brian Caulfield
共 18 条热搜查看完整榜单
纽约时报
中文网
10小时前更新
  • 01
    美国债务突破40万亿美元，借贷狂潮继续
    尽管美国仍是世界最大经济体，但不断攀升的债务负担可能导致投资者要求提高美国债券利率，或对美国信誉产生质疑，从而削弱对美元作为世界储备货币的信心。 Al Drago for The New York Times 特朗普总统曾承诺恢复财政秩序，但他旨在削减开支和增加收入的主要举措均未付诸实施。ALAN RAPPEPORT
  • 02
    郭德纲改编红歌被调查，中国言论空间日益收窄
    郭德纲因即兴改编一首红歌的歌词而遭到举报，目前正接受官方调查。随着习近平将文化艺术作为意识形态整顿的重点，中国演艺人士自我表达的空间日益收窄。 Shiho Fukada for The New York Times 2011年，郭德纲在北京表演相声。这种喜剧艺术形式融合了预先编排好的台词与现场即兴发挥。YAN ZHUANG
  • 03
    前福奇顾问承认试图隐瞒与新冠起源相关的记录
    莫伦斯承认合谋隐瞒与中国新冠病毒疫情起源有关的联邦记录。此前，共和党人指控福奇试图掩盖新冠病毒从中国实验室泄露的证据。莫伦斯的陈述并未指控福奇有任何不当行为。 Annabelle Gordon/Reuters 今年5月，前国立卫生研究院官员戴维·莫伦斯在马里兰州格林贝尔特联邦地区法院出庭后离开。BENJAMIN MUELLER
  • 04
    中国首次实现火箭陆地回收
    此次发射由民营初创企业蓝箭航天完成，此前中国已成功实现一子级海上平台回收。火箭一子级的重复使用将降低卫星制造商的成本，使中国公司在全球航天市场中更具竞争力。 Landscape/Xinhua, via Landscape/Xinhua Via ApADEEL HASSAN
  • 05
    朝鲜回应特朗普缩减美韩军演规模：“对我们毫无吸引力”
    金正恩胞妹金与正称特朗普的示好举动“对我们毫无吸引力”，并表示即使规模缩减，这些演习依然“具有挑衅性和侵略性”。她还表示并不了解美朝政府之间有任何沟通。 Jung Yeon-Je/Agence France-Presse — Getty Images 周三，在与美国举行的名为“乙支自由盾”的大型联合军演期间，韩国士兵在首尔一家酒店演习。ZANE IRWIN
  • 06
    中国机器人制造商宇树科技上市首日股价飙升500%
    宇树科技股价在周三早盘交易中上涨近500%，市值达到约530亿美元。宇树成为第二家受益于中国人工智能热潮的企业，此前内存芯片制造商长鑫上市首日股价飙升470%。 Qilai Shen for The New York Times 上周，在中国上海的一场行业展会上，参观者围聚在宇树科技的展位周围。STEVE LOHR
  • 07
    三位自以为是的领导人和他们的愚蠢战争
    特朗普、内塔尼亚胡和普京都自以为聪明，却犯下了最愚蠢的错误：他们都一头撞进了战争死胡同，如今只有咽下认输的苦果才能脱身。 Nikita Teryoshin托马斯·弗里德曼
  • 08
    攻击朋友、讨好敌人：特朗普“黑白颠倒”的世界观
    特朗普“攻击朋友、讨好敌人”的国际关系理念已成为当下美国外交政策的核心特征。美国与传统盟友之间的关系日渐疏远，而总统却频频向那些曾被视为对手的国家示好。 Doug Mills/The New York Times 特朗普总统那种“攻击朋友、讨好敌人”的做法已成为美国外交政策的常态。PETER BAKER
  • 09
    特朗普高强度赶工白宫宴会厅，力求“跑赢司法审查”
    特朗普在白宫建造宴会厅的计划正在最高法院待审。与此同时，他已招募一支由250名工人组成的团队，以每周七天、每天20小时的强度推进施工，以求在司法审查前完工。 Andrew Leyden for The New York Times 白宫管理与行政事务主任约书亚·费舍尔估计，白宫宴会厅的施工已完成65%。LUKE BROADWATER,ANN E. MARIMOW
  • 10
    朱镕基葬礼在北京举行，当局谨慎应对公众情绪
    当局不希望公众将朱镕基主政时代与现在进行对比。朱镕基遗体告别仪式前一天，习近平在江泽民诞辰百年纪念活动上强调了自己的执政与江泽民执政的延续性。 Maxim Shemetov/Reuters 周二，北京天安门广场为前国务院总理朱镕基降半旗致哀。YAN ZHUANG
  • 11
    如果北京武力攻台，普通人能“保卫台湾”吗？
    中国在台湾周边的军事演习日益频繁。越来越多的台湾民众参与民防演习，学习如何疏散伤员、提供急救和营救人质。但也有人批评民防倡导者是在制造恐慌、鼓吹战争。BROOK LARMER
  • 12
    战争和权力的新物理规律
    从乌克兰到伊朗，战争中相对弱势的一方如今能以极低成本、极高精度展开行动。决定战争胜负的曾经是质量和能量，如今则是信息。 Gregory Halpern for The New York Times托马斯·弗里德曼
  • 13
    为何美日联手也“救”不了日元？
    此前美国和日本花费数百亿美元短暂拉升日元，但目前日元再次下跌，抹去了干预所带来的部分涨幅，再次逼近160日元关口。这可能会考验东京和华盛顿官员的决心。 The Yomiuri Shimbun, via Associated Press 日元涨势已消退，汇率正逼近160日元兑1美元，这一水平可能会考验东京和华盛顿官员的决心。RIVER AKIRA DAVIS
  • 14
    在伊朗陷入困局后，特朗普再次将目光转向金正恩
    在伊朗问题上陷入僵局后，特朗普暗示自己急于结束这场冲突，转而处理朝鲜问题。但分析人士认为，金正恩现在对改善与美国的关系根本没有兴趣，朝鲜不可能放弃核武器。 Erin Schaff/The New York Times 2019年，特朗普总统与朝鲜领导人金正恩在朝韩非军事区会面。DAVID E. SANGER,ANTON TROIANOVSKI
  • 15
    中美人才争夺战：中国加速追赶，美国“自废武功”？
    在中美科技霸主竞争中，谁能吸引到顶尖人才至关重要。中国近年来大力投入资金招募人才，美国则因科研经费削减、限制性移民等政策削弱了其原有的优势。 Dongyan Xu袁莉
  • 16
    AI垃圾满天飞，科技公司开展人工智能废物大扫除
    互联网上充斥着人工智能产生的垃圾内容，使信息生态系统遭到污染。现在，科技公司开始正视这一问题，运用各种策略来扫除这些生成式废物，并动员用户帮助识别垃圾内容。 Lia Sued C.TIFFANY HSU
  • 17
    硅谷并没有“右转”，它本来就“右”
    硅谷与白宫或五角大楼的真正界限与意识形态关系不大，而是与市场必不可分。当它们无法从中国市场赚取利润，大公司就开始争夺国防业务。 Brendan ConroySHARON WEINBERGER
  • 18
    乌克兰无人机何以精准打击俄罗斯本土？
    太空已经成为俄乌战争的下一个前沿阵地。高科技卫星影像为乌克兰军方近乎实时地提供俄军的移动和位置情报，令无人机能够迅速打击俄占区的纵深腹地乃至俄罗斯本土。 上月在乌克兰扎波罗热地区，一名乌克兰士兵正在为一架无人机打击俄军目标做准备。MICHAEL SCHWIRTZ,TYLER HICKS
  • 19
    干预政策、压制批评，特朗普政府将美国签证“武器化”
    特朗普政府正利用美国签证政策来加强对拉丁美洲的控制，通过威胁撤销签证来向该地区的政治人物施压，迫使其接受自己的要求，或惩罚那些对美国政策提出质疑的人。 Kenny Holston/The New York Times 由鲁比奥领导的国务院已恢复了中美洲一些曾受美国政府制裁的政治人物的签证，同时也撤销了另外一些人的签证。EMILIANO RODRÍGUEZ MEGA,EMMA BUBOLA,GABRIEL LABRADOR,DAVID BOLAÑOS
  • 20
    具有中国特色的AI数据：中国争夺人工智能时代话语权
    北京计划在2028年底前将中国打造成数据强国，希望借此缩小与美国在获取高质量AI训练数据方面的差距，在人工智能聊天机器人开发中拥有更大的话语权。 Go Nakamura/Reuters 上个月在上海举行的世界人工智能大会上的谷歌展台。中国希望在人工智能聊天机器人的开发中拥有更大的话语权。DAVID PIERSON,BERRY WANG
共 20 条热搜查看完整榜单
One Useful Thing
实时热榜
13小时前更新
共 20 条热搜查看完整榜单
Open Robotics Blog
实时热榜
13小时前更新
  • 01
    OSRA announces technology strategy for 2026
    The OSRA is pleased to announce that the Open Robotics Technology Strategy for 2026 has been published.Geoffrey Biggs
  • 02
    OSRA Projects Documentation Overhaul: New Information Architecture Design Complete
    The OSRA is pleased to announce the completion of a significant milestone in overhauling our project documentation: the design of the new information architecture is complete, and we are ready to begin modifying our documentation tool chains and content to deploy it.Geoffrey Biggs
  • 03
    New Build Farm Backer Program & Infra Team Swag
    Ever wonder how sudo apt install just works for ROS? That’s the magic of the OSRF Build Farm! This beast serves over half a billion packages a year, saving us all from the headache of compiling from source. But maintaining this critical utility takes massive server power and a dedicated team of engineers. To keep the lights on and the builds green, we are launching the Build Farm Backer campaign! We need your help to support the hardware and the humans behind the open source robotics ecosystem.Open Robotics
  • 04
    Open Robotics launches Zulip chat server
    Please join us on the official Open Robotics Zulip chat server. This chat server brings together all parts of our community, including the project maintainers, in a unified space where everyone can talk to each other and see what’s going on.Geoffrey Biggs
  • 05
    Launch of the OSRA's new Robotics Enhancement Proposal (REP) process
    The OSRA Technical Governance Committee has officially ratified a new Robotics Enhancement Proposal (REP) process. This is the successor to the original ROS Enhancement Proposal process and creates a unified framework for developing technical specifications across all OSRA projects.Geoffrey Biggs
  • 06
    OSRA's Technical Governance Committee Approves $250,000 Funding for Infrastructure and Documentation Enhancement
    The Technical Governance Committee (TGC) of the Open Source Robotics Alliance (OSRA) has approved funding of US$250,000 to significantly enhance project infrastructure and documentation across the OSRA ecosystem. This funding will focus on maximising value across all OSRA projects through strategic improvements in two key areas.Geoffrey Biggs
  • 07
    Google Summer of Code 2025
    Meet our 2025 Google Summer of Code students!Katherine Scott
  • 08
    ROS 2 Kilted Kaiju Released
    Happy World Turtle Day! Today, we are happy to announce the eleventh release of ROS 2: Kilted Kaiju! Kilted is a standard release, and will be supported until November, 2026. If you like the fantasic release art by our artist, Ryan Hungerford, our Kilted t-shirt and merchandise store is still open! All proceeds help support the Open Source Robotics Foundation (OSRF) and its projects.Geoffrey Biggs
  • 09
    OSRF Adopts Policy on Use of Generative AI in Contributions
    The rise of Generative AI tools in software development presents both challenges and opportunities for open-source projects. While these tools can boost productivity and creativity, they also raise concerns about ownership and ethical use. The OSRF is developing a comprehensive policy to address these issues, ensuring project integrity and fostering an inclusive environment. The Technical Governance Committee initiated a Technical Committee to explore how other foundations handle these challengeGeoffrey Biggs
  • 10
    OPEN ROBOTICS ANNOUNCES SOLD-OUT ROSCON 2024
    ROSCon 2024 is sold out!Vanessa Yamzon Orsi
  • 11
    OSRF joins the Eclipse Foundation
    The Open Source Robotics Foundation (OSRF, Open Robotics) is very proud to announce that it has joined the Eclipse Foundation . The Eclipse Foundation is one of the premier open-source software foundations in the world. Originally created by IBM in 2001 to act as an unbiased steward of the Eclipse open-source IDE and its developer community, the Eclipse Foundation is now home to well over 400 open-source software projects. Users of OSRF software such as ROS may recognise it from ROS’s use of EclGeoffrey Biggs
  • 12
    Open-RMF gets Jazzy
    We're happy to announce the availability of the latest binary and source packages for Open-RMF on ROS 2 Jazzy Jalisco ! This is a feature packed release along with several critical bug fixes and improvements. The most notable features, available exclusively on the Jazzy release, are: Commissioning : The ability to selectively commission/decommission robots from particular fleets and manage how tasks are assigned to the robot during these phases. Runtime charging swaps : Allow robots to swap charYadu
  • 13
    Open Robotics and Google Summer of Code 2024
    It’s time for another Google Summer of Code (GSoC)! We have five contributors this year, all working on improving aspects of Gazebo. Rakesh Vivekanandan, Helena Moyan and Gaurav Kumar will be working with Woensug Choi and Mabel Zhang on physics-based sonar simulation for underwater robots. Saurabh Kamat will be working with Jose Luis Rivero on improving how users install Gazebo. Yaswanth Gonna is going to help Steve Peters put together an improved set of worlds for benchmarking physics engines.Geoffrey Biggs
  • 14
    ROS 2 Jazzy Jalisco Released
    The OSRF is are happy to announce the tenth release of ROS 2, Jazzy Jalisco, released on May the 23rd, 2024! Jazzy is a Long Term Support release that will be supported until the end of May, 2029.Geoffrey Biggs
  • 15
    Announcing the Open Source Robotics Alliance
    Open Robotics (OSRF) is pleased to announce the creation of the Open Source Robotics Alliance (OSRA). The OSRA is a new initiative from the OSRF to ensure the long-term stability and health of our open-source robot software projects.Open Robotics
  • 16
    Audrow Nash to step down from hosting the Sense, Think, Act podcast
    Audrow Nash is moving on to the next phase of his podcasting career and will be stepping down as the host of the OSRF’s Sense, Think, Act podcast. Under Audrow’s masterful leadership and hosting, Sense, Think, Act has become one of the premier places to hear interviews with leaders in the world of robotics.Open Robotics
  • 17
    ROSCon 2023 Recap
    ROSCon 2023, our twelfth annual ROS developers' conference, was held in New Orleans, Louisiana on October 18-20, 2023. The conference was a success, bringing together over 700 ROS developers from 42 countries to learn from each other and share the latest advances in ROS and robotics technology. We were also thrilled to see the Dronecode Foundation’s flagship PX4 Developers’ Summit hosted immediately after ROSCon in the same venue, allowing a number of attendees to attend both events and learn frOpen Robotics
  • 18
    Gazebo Harmonic Released!
    The Gazebo team is happy to announce the 8th release of Gazebo, “Gazebo Harmonic”. This new release represents a year of development by our Gazebo developer community, and it is one of our biggest Gazebo releases yet! The release is jam packed full of new features, and we believe that most Gazebo users will be excited to upgrade as soon as possible! A complete list of features can be found in the release notes, but we wanted to take a moment to point out some of the highlights.Open Robotics
  • 19
    Google Summer of Code Virtual RobotX
    Meet Tejal, one of our Google Summer of Code students for 2023.Open Robotics
  • 20
    Gazebo 2023 Google Summer of Code
    We’re really excited to announce our Google Summer of Code (GSoC) students for 2023. This year we are mentoring two GSoC students who are helping us with the development of new features for Gazebo that will hopefully make it into the Gazebo Harmonic release later this year. We asked the students to share a little bit about themselves and what they plan to work on over the summer.Open Robotics
共 20 条热搜查看完整榜单
OpenAI
News
15天前更新
共 50 条热搜查看完整榜单
OpenAlex
实时热榜
8小时前更新
共 50 条热搜查看完整榜单
OpenReview
实时热榜
7小时前更新
  • 01
    TimePre: Bridging Accuracy, Efficiency, and Stability in Probabilistic Time-Series Forecasting
    期刊：Transactions on Machine Learning Research · 摘要：We propose TimePre, a simple framework that unifies the efficiency of Multilayer Perceptron (MLP)-based models with the distributional flexibility of Multiple Choice Learning (MCL) for Probabilistic Time-Series Forecasting (PTSF). Stabilized Instance Normalization (SIN), the core of TimePre, is a normalization layer that explicitly addresses the trade-off among accuracy, efficiency, and stability. SIN stabilizes the hybrid architecture by correcting channel-wise statistical shifts, thereby preventing the hypothesis collapse that otherwise destabilizes this combination. Extensive experiments o… · 篇幅：Regular submission (no more than 12 pages of main content) · 代码：https://github.com/LyCharles/TimePre/ · OpenReview ID：yQLnvkJMbPLingyu Jiang, Lingyu Xu, Peiran Li et al.
  • 02
    When Vision Needs a Second Look: Tool-Augmented Active Perception for Earth Observation
    期刊：Transactions on Machine Learning Research · 摘要：Earth Observation (EO) uses satellite and aerial imagery to monitor the Earth’s surface, supporting critical applications in infrastructure, agriculture, and climate change. As governments and industry scale EO pipelines, reliable automation has become essential. Yet, current Vision-Language Models are limited to coarse-grained perception, struggling to execute the precise, multi-step reasoning required for operational decision-making. Recent evaluations on benchmarks like GeoBench-VLM highlight this shortcoming: even state-of-the-art models show low accuracy and frequently struggle with task… · 篇幅：Regular submission (no more than 12 pages of main content) · OpenReview ID：7yUrnyFgEqAtul Dev, Shivank Garg, Gaurav Kumar Nayak
  • 03
    Gradient-Based Multi-Objective Deep Learning: Algorithms, Theories, Applications, and Beyond
    期刊：Transactions on Machine Learning Research · 摘要：Many modern deep learning applications require balancing multiple objectives that are often conflicting. Examples include multi-task learning, fairness-aware learning, and the alignment of Large Language Models (LLMs). This leads to multi-objective deep learning, which tries to find optimal trade-offs or Pareto-optimal solutions by adapting mathematical principles from the field of Multi-Objective Optimization (MOO). However, directly applying gradient-based MOO techniques to deep neural networks presents unique challenges, including high computational costs, optimization instability, and the… · 篇幅：Long submission (more than 12 pages of main content) · OpenReview ID：eCUcXXH3PSWeiyu Chen, Baijiong Lin, Xiaoyuan Zhang et al.
  • 04
    Efficient Image Restoration with State-Dependent Forward Diffusion
    期刊：Transactions on Machine Learning Research · 摘要：This paper proposes to perform image restoration through a state-dependent mean-reverting forward diffusion (FoD) process. In contrast to traditional diffusion-based approaches that rely on a coupled forward-backward diffusion scheme, FoD directly learns image restoration through a single forward diffusion process, yielding a simple yet efficient framework. The core of FoD is a state-dependent stochastic differential equation (SDE) that involves a mean-reverting term in both the drift and diffusion functions. This mean-reverting structure drives the low-quality data toward the clean endpoint… · 篇幅：Regular submission (no more than 12 pages of main content) · 代码：https://github.com/Algolzw/FoD · OpenReview ID：Eq9k6Va3hYZiwei Luo, Fredrik K. Gustafsson, Jens Sjölund et al.
  • 05
    TreeSMOTE: Structure-Aware Data Augmentation for Imbalanced Tabular Learning
    期刊：Transactions on Machine Learning Research · 摘要：Class imbalance has been a critical bottleneck in classification problems, undermining a classifier's identification of minority instances. Data augmentation provides an effective solution by oversampling the minority. Extant methods often generate samples through duplication, perturbation, or interpolation, largely relying on the assumption of local smoothness of the data space to ensure synthetic data reliability. Alternatively, generative models are leveraged for data learning and synthesis. However, both approaches encounter significant limitations in tabular data, primarily due to data h… · 篇幅：Regular submission (no more than 12 pages of main content) · 代码：https://github.com/1836897243/TreeSmote · OpenReview ID：OelOS8cbBYPeng Wang, Hangting Ye, He Zhao et al.
  • 06
    WorldPack: Dynamic Frame Compression for Long-context Video World Modeling
    期刊：Transactions on Machine Learning Research · 摘要：Video world models have attracted significant attention for their ability to produce high-fidelity future visual observations conditioned on past observations and navigation actions. However, achieving temporally and spatially consistent generation over long horizons remains an open challenge: existing approaches either compress past frames at fixed rates based on temporal proximity, discarding spatially critical information, or retrieve only a handful of relevant frames without increasing the total amount of retained history. In this paper, we propose WorldPack, a video world model that intr… · 篇幅：Regular submission (no more than 12 pages of main content) · OpenReview ID：zJuiG3PiNJYuta Oshima, Yusuke Iwasawa, Masahiro Suzuki et al.
  • 07
    TS-Reasoner: Aligning Time Series Foundation Models with LLM Reasoning
    期刊：Transactions on Machine Learning Research · 摘要：Time series reasoning is crucial to decision-making in diverse domains, including finance, energy, and scientific discovery. While existing time series foundation models (TSFMs) can capture low-level dynamic patterns and provide accurate forecasting, further analysis usually requires additional background knowledge and sophisticated reasoning, which are lacking in most TSFMs but can be achieved through Large Language Models (LLMs). On the other hand, without expensive post-training, LLMs often struggle with the numerical understanding of time series data. Although it is intuitive to integrate… · 篇幅：Regular submission (no more than 12 pages of main content) · OpenReview ID：d6TD0f2xXqFangxu Yu, Hongyu Zhao, Tianyi Zhou
  • 08
    A Survey on Hallucination in Video Understanding: Taxonomy, Causes, and Mitigation Techniques
    期刊：Transactions on Machine Learning Research · 摘要：Video Large Language Models (Vid-LLMs) have recently achieved strong performance across a wide range of video understanding tasks, including question answering, captioning, and multimodal reasoning. However, these models frequently produce outputs that are not faithfully grounded in the underlying video content, a phenomenon commonly referred to as hallucination. Compared with hallucination in text-only or image-based models, hallucination in video understanding is further complicated by temporal dynamics, motion interpretation, long-context dependencies, and event-level reasoning. In this su… · 篇幅：Long submission (more than 12 pages of main content) · OpenReview ID：qbO71rVrIGJiayi Sheng, Wei Luo, Wotao Yin
  • 09
    Exploiting Completeness Perception with Diffusion Transformer for Unified 3D MRI Synthesis
    期刊：Transactions on Machine Learning Research · 摘要：Missing data problems, such as missing modalities in multi-modal brain MRI and missing slices in cardiac MRI, pose significant challenges in clinical practice. Existing methods rely on external guidance to supply detailed missing-state information for instructing generative models to synthesize missing MRIs. However, manual indicators are not always available or reliable in real-world scenarios due to the unpredictable nature of clinical environments. Moreover, these explicit masks are not informative enough to provide guidance for improving semantic consistency. In this work, we argue that g… · 篇幅：Regular submission (no more than 12 pages of main content) · 代码：https://github.com/JK-Liu7/CoPeDiT · OpenReview ID：DCaolE9oBNJunkai Liu, Nay Aung, Theodoros N. Arvanitis et al.
  • 10
    From Uniform to Learned Knots: A Study of Spline-Based Numerical Encodings for Tabular Deep Learning
    期刊：Transactions on Machine Learning Research · 摘要：Numerical preprocessing remains a critical component of tabular deep learning, as the representation of continuous features can strongly affect downstream performance. We systematically study spline-based numerical encodings, including B-splines, M-splines, and integrated splines (I-splines), under uniform, quantile-based, target-aware, and learnable-knot placement. For the learnable variants, we adopt a differentiable knot parameterization that enables stable end-to-end optimization of knot locations jointly with the backbone. We evaluate these encodings on a diverse collection of public reg… · 篇幅：Long submission (more than 12 pages of main content) · 代码：https://github.com/mkumar73/tdl-numerical-encodings/ · OpenReview ID：str7wQt9QcManish Kumar, Anton Frederik Thielmann, Christoph Weisser et al.
  • 11
    SimpleDesign: A Joint Model for Protein Sequence and Structure Codesign
    期刊：Transactions on Machine Learning Research · 摘要：Proteins are fundamental to biological processes, with their function determined by the complex interplay between the amino acid sequence and the three-dimensional structure. Developing generative models capable of understanding this intrinsically multi-modal relationship is crucial for fields like drug discovery and protein engineering. Existing models often rely on a multi-stage training process where autoencoders that tokenize data into latent representations are trained in a first stage. Secondly, a generative model is trained on the latent representation of the autoencoder(s), i.e., gene… · 篇幅：Regular submission (no more than 12 pages of main content) · OpenReview ID：wPfw7GkMnsJiarui Lu, Yuyang Wang, Yizhe Zhang et al.
  • 12
    LaCy: What Small Language Models Can and Should Learn is Not Just a Question of Loss
    期刊：Transactions on Machine Learning Research · 摘要：Language models have consistently grown to compress more world knowledge into their parameters, but the knowledge that can be pretrained into them is upper-bounded by their parameter size. Especially the capacity of Small Language Models (SLMs) is limited, leading to factually incorrect generations. This problem is often mitigated by giving the SLM access to an outside source: the ability to query a larger model, documents, or a database. Under this setting, we study the fundamental question of \emph{which tokens an SLM can and should learn} during pretraining, versus \emph{which ones it shou… · 篇幅：Regular submission (no more than 12 pages of main content) · OpenReview ID：JxVxBa3wO5Szilvia Ujváry, Louis Béthune, Pierre Ablin et al.
  • 13
    Understanding and Mitigating Overconfidence in Focus Group Surveys
    期刊：Transactions on Machine Learning Research · 摘要：Subjective evaluation tasks including critical analysis and rating remain at the top of Bloom’s Taxonomy. These have emerged as new pathways for evaluating Language Models (LMs) wherein correctness is relative. While LMs present diverse and human-aligned opinions on such tasks, their confidence and reliability in opinions remains unexplored. We take a deeper look at the reliability of LMs for subjective evaluations by selecting one such task of focus group surveys. LMs act as participants by completing survey questionnaires of diverse physical products. Participants must verbalize their opini… · 篇幅：Regular submission (no more than 12 pages of main content) · 代码：https://github.com/karush17/focus-groups · OpenReview ID：NGuOZYQZBqKarush Suri, Konstantinos N. Plataniotis, Yuri Andrew Lawryshyn
  • 14
    CURE-OOD: Benchmarking Out-of-Distribution Detection for Survival Prediction
    期刊：Transactions on Machine Learning Research · 摘要：``How long can I live and remain free of cancer?'' is often the first question a patient asks after receiving a cancer diagnosis and treatment. Accurate survival prediction helps alleviate psychological distress and supports risk stratification and personalized treatment planning. Recent survival prediction frameworks have shown strong performance using computed tomography (CT) images. However, variations in imaging acquisition introduce out-of-distribution (OOD) samples caused by covariate shifts that undermine model reliability. Despite this challenge, to our knowledge, no existing benchmar… · 篇幅：Regular submission (no more than 12 pages of main content) · 代码：https://github.com/WenjieZhao1/CURE-OOD · OpenReview ID：fziI7nE1vOWenjie Zhao, Jia Li, Mingrui Liu et al.
  • 15
    Hierarchical Diffusion for Efficient and Transferable Climate Downscaling
    期刊：Transactions on Machine Learning Research · 摘要：Downscaling is essential for generating the high-resolution climate data needed for local planning, but traditional methods remain computationally demanding. Recent years have seen impressive results from AI downscaling models, particularly diffusion models, which have attracted attention due to their ability to generate ensembles and overcome the smoothing problem common in other AI methods. However, these models typically remain computationally intensive. We introduce a Hierarchical Diffusion Downscaling (HDD) model, which introduces an easily-extensible hierarchical sampling process to the… · 篇幅：Regular submission (no more than 12 pages of main content) · OpenReview ID：OhTYgFpMU2Declan Jacques Curran, Sanaa Hobeichi, Hira Saleem et al.
  • 16
    Gradient Heterogeneity Complements Hessian Heterogeneity in Transformer Optimization
    期刊：Transactions on Machine Learning Research · 摘要：Transformers are difficult to optimize with stochastic gradient descent (SGD) and largely rely on adaptive optimizers such as Adam. Despite extensive efforts, the mechanisms behind Adam's advantage over SGD in Transformer optimization are still not fully understood. In this study, we analyze the optimization of Transformer models in the fine-tuning setting through the lens of gradient heterogeneity, defined as the variation in gradient norms across parameter blocks. We provide a theoretical analysis showing that gradient heterogeneity, together with Hessian heterogeneity, degrades the converg… · 篇幅：Regular submission (no more than 12 pages of main content) · 代码：https://github.com/tom4649/gradient-heterogeneity · OpenReview ID：wZJcQb5m1eAkiyoshi Tomihari, Issei Sato
  • 17
    Efficient LLM Collaboration via Planning
    期刊：Transactions on Machine Learning Research · 摘要：Recently, large language models (LLMs) have demonstrated strong performance, ranging from simple to complex tasks. However, while large models achieve remarkable results across diverse tasks, they often incur substantial monetary inference cost, making frequent use impractical for many applications. In contrast, small models are often freely available and easy to deploy locally, but their performance on complex tasks remains limited. This trade-off raises a natural question: how can small and large models efficiently collaborate to combine their complementary strengths? To bridge this trade-o… · 篇幅：Regular submission (no more than 12 pages of main content) · 代码：https://github.com/prinwinter/cope · OpenReview ID：RPzbeL0koPByeongchan Lee, Jonghoon Lee, Dongyoung Kim et al.
  • 18
    Compress--Add--Smooth: Fixed-Budget Temporal Compression of Density-Valued Streams
    期刊：Transactions on Machine Learning Research · 摘要：We study fixed-budget temporal memory for streams of probability distributions. The proposed representation is a piecewise-linear density protocol on a replay interval $[0,1]$: recent experience is stored near $t=1$, older experience is represented by intermediate-time marginals, and new experience is incorporated by a deterministic \emph{Compress--Add--Smooth} (CAS) recursion. In the Gaussian-mixture instantiation considered here, each protocol node stores a labeled $K$-component Gaussian mixture in $d$ dimensions, and each daily update costs $O(LKd^2)$ arithmetic operations for a fixed temp… · 篇幅：Long submission (more than 12 pages of main content) · 代码：https://github.com/mchertkov/compress-add-smooth · OpenReview ID：wjoixYG0mCMichael Chertkov
  • 19
    A Benchmark for Vericoding: Formally Verified Program Synthesis
    期刊：Transactions on Machine Learning Research · 摘要：We present and test the largest benchmark for vericoding, LLM-generation of formally verified code from formal specifications — in contrast to vibe coding, which generates potentially buggy code from a natural language description. Our benchmark contains 12,504 formal specifications, with 3,029 in Dafny, 2,334 in Verus/Rust and 7,141 in Lean. Of these, 6,174 are new unseen problems. We find vericoding success rates of 27% in Lean, 44% in Verus/Rust and 82% in Dafny using off-the-shelf LLMs. Adding natural-language descriptions does not significantly improve performance. We also find that LLM… · 篇幅：Regular submission (no more than 12 pages of main content) · OpenReview ID：Zgh5kpGAm8Ionel Emilian Chiosa, Sergiu Bursuc, Theodore Ehrenborg et al.
  • 20
    Dynamic Subspace Estimation from Undersampled Data using Grassmannian Geodesics
    期刊：Transactions on Machine Learning Research · 摘要：This work considers recovering a sequence of low-rank matrices from undersampled measurements, where the underlying subspace varies across samples over time. Existing works involve concatenating all of the samples from each time point to recover the underlying matrix under the assumption that the data are well-approximated by a single, static subspace. However, this assumption is inappropriate for applications where the best low-rank approximations vary over time. To address this issue, we propose a Riemannian block majorize minimization algorithm that constrains the time-varying subspaces as… · 篇幅：Long submission (more than 12 pages of main content) · OpenReview ID：W0ZNwXoBehSoo Min Kwon, Cameron James Blocker, Haroon Raja et al.
  • 21
    Yose-Ue: A Treap-Based Ensemble Framework for Resource-Efficient Unsupervised Anomaly Detection
    期刊：Transactions on Machine Learning Research · 摘要：Anomaly detection seeks to identify observations that deviate significantly from an underlying data distribution. While deep learning and ensemble-based approaches have achieved strong empirical performance, their computational and memory requirements limit their applicability in resource-constrained edge environments. Furthermore, many approaches to improving efficiency rely on supervised models, which require labeled anomalies that are often scarce in practice. We propose Yose-Ue, a resource-efficient, fully unsupervised anomaly detection framework based on treap-structured ensemble learnin… · 篇幅：Long submission (more than 12 pages of main content) · 代码：https://github.com/eduardoortegathethird/Yose-Ue · OpenReview ID：17y2ooyemGEduardo Ortega, Gautam Dasarathy, Krishnendu Chakrabarty
  • 22
    A Robust $\widetilde{\mathcal{O}}(1/\sqrt{T})$ Rate for Unprojected TD Learning with Linear Function Approximation
    期刊：Transactions on Machine Learning Research · 摘要：We investigate the finite-time convergence properties of Temporal Difference (TD) learning with linear function approximation, a cornerstone of reinforcement learning. We are interested in the so-called ``robust'' setting, where the convergence guarantee does not depend on the potential function's minimal curvature. While prior work has established convergence guarantees in this setting, these results typically rely on the artificial assumption that each iterate is projected onto a bounded set. Removing such a condition was left as an open problem by Bhandari et al. (COLT'18), hypothesizing t… · 篇幅：Regular submission (no more than 12 pages of main content) · OpenReview ID：Tj1B5WDyt8Wei-Cheng Lee, Francesco Orabona
  • 23
    Density-Scaled Regularization for Offline Reinforcement Learning
    期刊：Transactions on Machine Learning Research · 摘要：Value-based offline RL methods are prone to overestimate the values of out-of-distribution (OOD) actions, and this is often addressed by regularizing the action-value function in the Bellman update. However, existing regularization methods can suffer from being too conservative, which can arise from over-penalizing the values for both in-distribution actions and out-of-support actions. We present a new regularization method for offline value-based methods, called Density-Scaled (DS) regularization, which penalizes the value function based on the relative action density of the behavior policy.… · 篇幅：Regular submission (no more than 12 pages of main content) · 代码：https://github.com/jackyxie5/density-scaled-regularization · OpenReview ID：nDPrzkQTj1Jacky Xie, Nan Ye
  • 24
    Adjoint Matching through the Lens of the Stochastic Maximum Principle in Optimal Control
    期刊：Transactions on Machine Learning Research · 摘要：Reward fine-tuning of diffusion and flow models and sampling from tilted or Boltzmann distributions can both be formulated as stochastic optimal control (SOC) problems, where learning an optimal generative dynamics corresponds to optimizing a control under SDE constraints. In this work, we revisit and generalize \emph{Adjoint Matching}, a recently proposed SOC-based method for learning optimal controls, and place it on a rigorous footing by deriving it from the \emph{Stochastic Maximum Principle} (SMP). We formulate a general Hamiltonian adjoint matching objective for SOC problems with contro… · 篇幅：Regular submission (no more than 12 pages of main content) · 代码：https://github.com/frankhan91/smp-adjoint-matching · OpenReview ID：tR5VsdQFhKCarles Domingo-Enrich, Jiequn Han
  • 25
    Benchmarking Transfer Learning: From Simple Baselines to Combined Scorers for Transferability Estimation
    期刊：Transactions on Machine Learning Research · 摘要：In the evolving landscape of deep learning, selecting the best pre-trained models from a growing number of choices is a challenge. Transferability scorers propose an efficient alternative to this challenge by calculating a proxy to rank a pool of pre-trained model candidates. Despite their promise, the field currently lacks standardized evaluation protocols, consistent baselines, and reproducible methodologies. This has led to contradictory findings across studies, with the best scorer in one study ranking among the worst in another. In this work, we introduce a benchmark for transferability… · 篇幅：Regular submission (no more than 12 pages of main content) · 代码：https://github.com/VirtualSpaceman/tmlr-transferability-benchmark · OpenReview ID：3i2ZRk8GDNLevy Chaves, Claudio Mayrink Verdun, Eduardo Valle et al.
  • 26
    FIMP: Foundation Model-Informed Message Passing for Graph Neural Networks
    期刊：Transactions on Machine Learning Research · 摘要：Foundation models have achieved remarkable success across many domains, relying on pretraining over vast amounts of data. Graph-structured data often lacks the same scale as unstructured data, making the development of graph foundation models challenging. In this work, we propose Foundation-Informed Message Passing (FIMP), a message-passing framework that repurposes existing pretrained non-textual foundation models for graph-based tasks in multiple capacities, including tokenization, representation extraction, and weight initialization. We show that the self-attention layers of foundation mod… · 篇幅：Regular submission (no more than 12 pages of main content) · OpenReview ID：fj7sjOwtXcSyed A Rizvi, Nazreen Pallikkavaliyaveetil MohammedSheriff, David Zhang et al.
  • 27
    Detecting Distributional Treatment Responders with False Discovery Rate Control
    期刊：Transactions on Machine Learning Research · 摘要：In this paper, In this paper, we introduce causal responder detection (CARD), a method for distributionalwe introduce causal responder detection (CARD), a method for distributional responder analysis that identifies treated subjects whose outcomes significantly depart from the control response distribution while controlling the false discovery rate (FDR) marginally over the tested treated population. CARD builds on the AdaDetect framework and, in randomized settings, inherits finite-sample FDR control under the exchangeability conditions required by AdaDetect when coupled with the Benjamini–H… · 篇幅：Regular submission (no more than 12 pages of main content) · OpenReview ID：zPt0o32mYnTzviel Frostig, Bradley P. Carlin, Elad Berkman et al.
  • 28
    EFFEKT: Efficient Federated Knowledge Transfer to Foundation Models
    期刊：Transactions on Machine Learning Research · 摘要：Recent data protection laws have accelerated the adoption of Federated Learning (FL) for privacy-preserving decentralized training. Nevertheless, increasing model sizes impose substantial computational demands on client devices, limiting FL applicability in resource-constrained settings. We introduce a novel multi-domain federated learning framework in which lightweight client-side proxy models collaborate with a server-side Foundation Model (FM) to learn new concepts without sharing private data. Our approach, EFFEKT, enables efficient server-side training of domain-specific LoRA adapters wh… · 篇幅：Regular submission (no more than 12 pages of main content) · 代码：https://github.com/LTTM/EFFEKT · OpenReview ID：jpUDUJfE1KMatteo Caligiuri, Francesco Barbato, Pietro Zanuttigh et al.
  • 29
    Optimization as a Dynamical System: Generative Schedules from Latent ODEs
    期刊：Transactions on Machine Learning Research · 摘要：We present a new meta-learning method to determine the optimal learning rate schedule for gradient descent. It leverages training runs from a hyperparameter search to learn a latent representation of the training process, which is modeled as a dynamical systems. Given current training metrics, it predicts the future learning rate schedule with the best long-term validation performance. Our scheduler generalizes beyond previously observed training dynamics and creates specialized schedules that deviate noticeably from even the best-performing parametric functions. It outperforms all baselines… · 篇幅：Regular submission (no more than 12 pages of main content) · OpenReview ID：SwmzJgB9TAMatt L. Sampson, Peter Melchior
  • 30
    ImpMIA: Leveraging Implicit Bias for Membership Inference Attack
    期刊：Transactions on Machine Learning Research · 摘要：Determining which data samples were used to train a model, known as Membership Inference Attack (MIA), is a well-studied and important problem with implications on data privacy. SotA methods (which are black-box attacks) rely on training many auxiliary reference models to imitate the behavior of the attacked model. As such, they rely on assumptions which rarely hold in real-world settings: (i) the attacker knows the training hyperparameters; (ii) all available non-training samples come from the same distribution as the training data; and (iii) the fraction of training data in the evaluation s… · 篇幅：Regular submission (no more than 12 pages of main content) · 代码：https://github.com/yuvalgol123/ImpMIA-code · OpenReview ID：34bnVED6EZYuval Golbari, navve wasserman, Gal Vardi et al.
  • 31
    Riemannian t-SNE on Several Matrix Manifolds
    期刊：Transactions on Machine Learning Research · 摘要：Matrix manifolds play a fundamental role in machine learning, underpinning data representations (\textit{e.g.}, linear subspaces and covariance matrices) and optimization procedures. These manifolds follow Riemannian geometry, where intrinsic geometric structure plays an important role in geometric learning algorithms. However, traditional visualization methods based on Euclidean assumptions often fail to respect such non-Euclidean structure, leading to distortions in the resulting embeddings. To address this limitation, building upon the established Riemannian t-SNE paradigm, we develop thre… · 篇幅：Long submission (more than 12 pages of main content) · 代码：https://github.com/paradox-going/ManiReduce · OpenReview ID：4EZeC0JwqMRui Wang, Bin Shi, Chen Hu et al.
  • 32
    Centroid-Referenced Mahalanobis Matching (CRM): A Scalable, Representation-Based Framework for Causal Inference in Large Observational Studies
    期刊：Transactions on Machine Learning Research · 摘要：Matching for causal inference can be computationally expensive at scale and can silently change the target population when overlap is limited. We propose Centroid-Referenced Mahalanobis Matching (CRM), which replaces global pairwise search with stratified sampling in two reference coordinates: each unit's Mahalanobis distance from the treated centroid and its Fisher coordinate along the treated-control mean shift. All covariates enter through the treated covariance geometry; CRM is therefore not principal-component preprocessing followed by nearest-neighbor matching. For $n$ units and $p$ pre… · 篇幅：Long submission (more than 12 pages of main content) · 代码：https://github.com/KemingHu-D/crm-matching · OpenReview ID：z74epfCe3AKeming Hu, Yingpei He
  • 33
    FARM: Enhancing Molecular Representations with Functional Group Awareness
    期刊：Transactions on Machine Learning Research · 摘要：We introduce Functional Group-Aware Representations for Small Molecules (FARM), a novel foundation model designed to bridge the gap between SMILES (a linear string representation of molecular structures), natural language, and molecular graphs. The key innovation of FARM lies in its functional group (FG) annotation at the atomic level, which enables both FG-enhanced SMILES and FG graphs: SMILES are enriched with FG information to specify which functional group each atom belongs to, while the FG graph captures the molecular backbone by showing how the functional groups are connected. This toke… · 篇幅：Regular submission (no more than 12 pages of main content) · 代码：https://github.com/thaonguyen217/farm_molecular_representation · OpenReview ID：2All12TFlhThao Nguyen, Kuan-Hao Huang, Ge Liu et al.
  • 34
    Revealing Positive and Negative Role Models to Help People Make Good Decisions
    期刊：Transactions on Machine Learning Research · 摘要：We consider a setting where agents take action by following their role models in a social network, and study strategies for a social planner to help agents by revealing whether the role models are positive or negative. Specifically, agents observe a local neighborhood of possible role models they can emulate, but do not know their true labels. Revealing a positive label encourages emulation, while revealing a negative one redirects agents toward alternative options. The social planner observes all labels, but operates under a limited disclosure budget that it selectively allocates to maximize… · 篇幅：Regular submission (no more than 12 pages of main content) · 代码：https://github.com/knaggita/InformationDisclosure · OpenReview ID：jdcXfoENf0Avrim Blum, Keziah Naggita, Matthew Walter et al.
  • 35
    Missing Value Uncertainty: Could Collecting Missing Values Change the Prediction?
    期刊：Transactions on Machine Learning Research · 摘要：In mission critical domains such as sensor networks, operators often face the critical decision of whether to act on incomplete information or whether collecting missing values is likely to change the prediction. Existing methods typically focus on imputing missing values or quantifying model uncertainty, but they do not directly assess the stability of a prediction if missing values were to be revealed. To address this gap, we first introduce a framework for Missing Value Uncertainty (MVU), which is the distribution of predictions induced by incomplete inputs at inference time. We formalize… · 篇幅：Regular submission (no more than 12 pages of main content) · 代码：https://github.com/inouye-lab/MissingValueUncertainty/releases/tag/tmlr · OpenReview ID：BRWTS5e03ZDavid J. Burnett, Surojit Ganguli, Lance M. Kaplan et al.
  • 36
    OTIS: Learning High-Quality Time Series Features With Tiny Encoders
    期刊：Transactions on Machine Learning Research · 摘要：We introduce \texttt{OTIS}, an \textbf{o}pen \textbf{ti}me \textbf{s}eries encoder that yields high-quality time series features for downstream deployment on \emph{any} system, including resource-constrained wearables and industrial sensors. Currently, the development of powerful general-purpose encoders relies on the scaling laws hypothesis, using large encoder sizes to memorise the heterogeneous distributions of multi-domain training data. However, this reliance on scale creates a barrier to real-world utility, rendering deployment on resource-constrained systems infeasible due to strict me… · 篇幅：Regular submission (no more than 12 pages of main content) · 代码：https://github.com/oetu/otis · OpenReview ID：WW206A1TruÖzgün Turgut, Philip Müller, Martin J. Menten et al.
  • 37
    Speedrunning ImageNet Diffusion
    期刊：Transactions on Machine Learning Research · 摘要：Recent advances have significantly improved the training efficiency of diffusion transformers. However, these techniques have largely been studied in isolation, leaving unexplored the potential synergies from combining multiple approaches. We present SR-DiT (Speedrun Diffusion Transformer), a framework that systematically integrates token routing, architectural improvements, and training modifications on top of representation alignment. Our approach achieves FID 3.14 and KDD 0.290 on ImageNet-256 using only a 140M parameter model at 400K iterations without classifier-free guidance---comparabl… · 篇幅：Regular submission (no more than 12 pages of main content) · 代码：https://github.com/SwayStar123/SpeedrunDiT · OpenReview ID：0mYu3uPM3jSwayam Bhanded
  • 38
    Semantic F1 Scores: Fair Evaluation Under Fuzzy Class Boundaries
    期刊：Transactions on Machine Learning Research · 摘要：We propose Semantic F1 Scores, novel evaluation metrics for subjective or fuzzy multi-label classification that quantify semantic relatedness between predicted and gold labels. Unlike the conventional F1 metrics that treat semantically related predictions as complete failures, Semantic F1 incorporates a label similarity matrix to compute soft precision-like and recall-like scores, from which the Semantic F1 scores are derived. Unlike existing similarity-based metrics, our novel two-step precision-recall formulation enables the comparison of label sets of arbitrary sizes without discarding lab… · 篇幅：Regular submission (no more than 12 pages of main content) · 代码：https://github.com/gchochla/semantic-f1-score · OpenReview ID：U0YJpGuFEcGeorgios Chochlakis, Jackson Trager, Vedant Vipul Jhaveri et al.
  • 39
    FairNVT: Fair Classification via Noise Injection in Vision Transformers
    期刊：Transactions on Machine Learning Research · 摘要：This paper presents \textbf{FairNVT}, a lightweight debiasing framework for pretrained transformer-based encoders that improves prediction fairness while preserving task performance. FairNVT is motivated by the intuition that reducing sensitive-attribute information in the representation used by the downstream classifier can facilitate fairer predictions. Our approach learns task-relevant and sensitive embeddings via lightweight adapters, applies calibrated Gaussian noise to the sensitive embedding, and fuses it with the task representation. Together with orthogonality constraints and fairnes… · 篇幅：Regular submission (no more than 12 pages of main content) · OpenReview ID：rzm6gZrYglQiaoyue Tang, Sepidehsadat Hosseini, Mengyao Zhai et al.
  • 40
    Contractive MASO‑Generalized Predictors for Stable Latent‑Space Learning in JEPA
    期刊：Transactions on Machine Learning Research · 摘要：Joint Embedding Predictive Architectures (JEPAs) learn representations by predicting latent target embeddings from contextual views, but their predictors are typically shallow feed‑forward networks with limited control over multi‑step dynamics and stability. We introduce \emph{Learnable Iterated Function Systems} (LIFS), a recursive and contractive latent operator that replaces the standard JEPA predictor with a mixture of affine maps applied over multiple refinement steps. The mixture weights are conditioned on the context embedding, enabling input‑adaptive geometric structure while preservi… · 篇幅：Long submission (more than 12 pages of main content) · OpenReview ID：k2Z2gPOtlqBELLOULATA
  • 41
    RAWDet-7: A Multi-Scenario Benchmark for Object Detection and Description on Quantized RAW Images
    期刊：Transactions on Machine Learning Research · 摘要：Most vision models operate on 8-bit standard RGB (sRGB) images produced by dedicated image sensor processing pipelines designed for human perception rather than machine reasoning. In contrast, RAW images preserve sensor measurements, dynamic range, and fine-grained scene structure that can be critical for downstream understanding. Yet, progress in the RAW-domain vision remains limited by the lack of large-scale, high-quality benchmarks. To close this gap, we introduce RawDet-7, a multi-scenario benchmark for object detection and object description on quantized RAW images, comprising ~25k trai… · 篇幅：Regular submission (no more than 12 pages of main content) · OpenReview ID：UHTJrsYieoMishal Fatima, Shashank Agnihotri, Kanchana Vaishnavi Gandikota et al.
  • 42
    An Empirical Study into Clustering of Unseen Datasets with Self-Supervised Encoders
    期刊：Transactions on Machine Learning Research · 摘要：Can pretrained models generalize to new datasets without any retraining? We deploy pretrained image models on datasets they were not trained for, and investigate whether their embeddings form meaningful clusters. Our suite of benchmarking experiments uses encoders pretrained solely on ImageNet-1k with either supervised or self-supervised training techniques, deployed on image datasets that were not seen during training, and clustered with conventional clustering algorithms. This evaluation provides new insights into the embeddings of self-supervised models, which prioritize different features… · 篇幅：Regular submission (no more than 12 pages of main content) · 代码：https://github.com/scottclowe/zs-ssl-clustering/ · OpenReview ID：gdwg7ntmT5Scott C. Lowe, Joakim Bruslund Haurum, Sageev Oore et al.
  • 43
    Fast Sharpness-escaping Optimization for Long-tailed Learning
    期刊：Transactions on Machine Learning Research · 摘要：Deep neural networks often suffer from poor generalization in long-tailed settings. From a loss landscape perspective, this degradation is largely attributed to the tendency of the optimization process to converge into sharp, unstable minima for underrepresented data. We investigate the recently proposed Muon optimizer, providing theoretical evidence that its gradient orthogonalization balances deterministic update strength across positive-curvature modes, thereby reducing the relative dominance of sharp directions. While effective, the Muon optimizer imposes heavy computational overhead in l… · 篇幅：Regular submission (no more than 12 pages of main content) · OpenReview ID：I19SJW09znJiaan Luo, Ziyi Tang, Feng Hong et al.
  • 44
    A Comparative Study of Label-free Representation Quality Metrics in Deep Learning
    期刊：Transactions on Machine Learning Research · 摘要：We present a comparative study of label-free metrics for assessing the quality of representations in deep neural networks to understand their reliability under a wide variety of configurations. We group existing label-free metrics into three families based on their construction and analytically establish connections between metrics within the same family. We then characterise the sensitivity of spectral metrics through controlled synthetic experiments. Finally, all label-free metrics are evaluated against downstream task accuracy across a diverse set of 260 vision models on six datasets spann… · 篇幅：Long submission (more than 12 pages of main content) · 代码：https://github.com/dannyrichy/representation-quality-metrics · OpenReview ID：yknkAksqr1Daniel Richards Arputharaj, Gabriel Eilertsen, Daniel Jönsson
  • 45
    Private and interpretable clinical prediction with quantum-inspired tensor train models
    期刊：Transactions on Machine Learning Research · 摘要：Publicly available clinical machine learning models pose an underappreciated privacy risk: their parameters or outputs can be exploited to recover information from patients whose data were used during training. Moreover, this risk is exacerbated by models such as logistic regression (LR), which are typically preferred in clinical settings for their transparency. To assess this empirically, we attack LORIS, a publicly available LR model for immunotherapy response prediction hosted on a U.S. government website. From evaluations through its public interface, we recover the model parameters and i… · 篇幅：Long submission (more than 12 pages of main content) · 代码：https://github.com/joserapa98/tns4loris · OpenReview ID：QtG3fC1v5tJosé Ramón Pareja Monturiol, Juliette Sinnott, Roger G Melko et al.
  • 46
    Score-based Membership Inference on Diffusion Models
    期刊：Transactions on Machine Learning Research · 摘要：Membership inference attacks (MIAs) against Diffusion Models (DMs) raise pressing privacy concerns by revealing whether a sample was part of the training set. While existing methods typically rely on measuring reconstruction error across multiple denoising steps as a test statistic, they often incur significant computational overhead. In this work, we present a simple yet successful attack statistic using only the predicted noise vectors from the DM's denoiser, or equivalently, the score. Specifically, we show that the expected denoiser output points toward a kernel-weighted local mean of nea… · 篇幅：Regular submission (no more than 12 pages of main content) · 代码：https://github.com/mx-ethan-rao/SimA · OpenReview ID：Ckvsu5xRmfMingxing Rao, Bowen Qu, Daniel Moyer
  • 47
    MaskGT: Learning Task-Adaptive Connectivity in Graph Transformers
    期刊：Transactions on Machine Learning Research · 摘要：Graph Transformers (GTs) enable all-to-all interactions, but the optimal connectivity is task-dependent: some problems favor sparse, topology-aligned message passing, while others need global attention. We propose MaskGT, a GT-agnostic module that learns a discrete sparse gate over attention edges. By learning which node pairs may communicate within self-attention, MaskGT injects a task-adaptive relational inductive bias without fully committing to the input adjacency. Across synthetic and real-world benchmarks, MaskGT improves performance and robustness by suppressing spurious interactions u… · 篇幅：Regular submission (no more than 12 pages of main content) · 代码：https://github.com/anitasyang/mask-gt · OpenReview ID：CS4BJcbCGFAnita Yang, Ken-Ichi Kawarabayashi
  • 48
    Weaves, Wires, and Morphisms: Formalizing and Implementing the Algebra of Deep Learning
    期刊：Transactions on Machine Learning Research · 摘要：Despite deep learning models running well-defined mathematical functions, we lack a formal mathematical framework for describing model architectures. Ad-hoc notation, diagrams, and pseudocode poorly handle nonlinear broadcasting and the relationship between individual components and composed models. This paper introduces a categorical framework for deep learning models that formalizes broadcasting through the novel axis-stride and array-broadcasted categories. This allows the mathematical function underlying architectures to be precisely expressed and manipulated in a compositional manner. Th… · 篇幅：Long submission (more than 12 pages of main content) · 代码：https://github.com/mit-zardini-lab/pyncd · OpenReview ID：GiO8eom0jDVincent Abbott, Gioele Zardini
  • 49
    Generative Modeling with Bayesian Sample Inference
    期刊：Transactions on Machine Learning Research · 摘要：We present a novel view of diffusion-like generative modeling from the perspective of iterative Gaussian posterior inference. By treating the generated sample as an unknown variable, we formulate the sampling process in the language of Bayesian probability: at each step, a model predicts the unknown sample from our current belief state and we compute a posterior belief from that prediction. Based on this formulation, we propose the generative model Bayesian Sample Inference (BSI). In addition to a rigorous theoretical analysis, we show that our perspective includes Bayesian Flow Networks (BFN… · 篇幅：Regular submission (no more than 12 pages of main content) · 代码：https://github.com/martenlienen/bsi · OpenReview ID：n8qQPnalVWMarten Lienen, Marcel Kollovieh, Stephan Günnemann
  • 50
    KOALA: Koopman Operator Learning for WiFi-Based Anticipatory Human Motion Prediction
    期刊：Transactions on Machine Learning Research · 摘要：WiFi Channel State Information (CSI) has emerged as a privacy-preserving alternative to cameras for human pose estimation. However, existing approaches treat pose inference as an instantaneous regression problem and do not model temporal dynamics, making future motion prediction infeasible. Naively applying vision-based prediction methods compounds the estimation noise already present in CSI-derived poses, as autoregressive rollouts amplify errors at every step. We propose KOALA, the framework for human motion prediction directly from WiFi CSI, by lifting noisy CSI-derived pose sequences into… · 篇幅：Long submission (more than 12 pages of main content) · OpenReview ID：JQA0EfQIfjQuang-Anh N.D., Pham Minh Duc, Thao Pham Phuong et al.
共 50 条热搜查看完整榜单
OpenRouter
LLM 榜 · 本周
7小时前更新
共 100 条热搜查看更多
Packagist
实时热榜
7小时前更新
  • 019.2亿
    symfony/deprecation-contracts
    本周热度排名：1 · A generic function and convention to trigger deprecation notices · 累计下载：921,816,925 · 收藏：2,106
  • 0212.6亿
    symfony/polyfill-mbstring
    本周热度排名：2 · Symfony polyfill for the Mbstring extension · 累计下载：1,261,187,213 · 收藏：7,835
  • 0311.2亿
    psr/http-message
    本周热度排名：3 · Common interface for HTTP messages · 累计下载：1,117,829,161 · 收藏：7,130
  • 0410.9亿
    psr/container
    本周热度排名：4 · Common Container Interface (PHP FIG PSR-11) · 累计下载：1,094,623,382 · 收藏：10,040
  • 059.5亿
    symfony/service-contracts
    本周热度排名：5 · Generic abstractions related to writing services · 累计下载：948,305,918 · 收藏：2,612
  • 0611.8亿
    symfony/console
    本周热度排名：6 · Eases the creation of beautiful and testable command line interfaces · 累计下载：1,180,283,751 · 收藏：9,986
  • 0711.1亿
    guzzlehttp/psr7
    本周热度排名：7 · PSR-7 message implementation that also provides common utility methods · 累计下载：1,107,582,214 · 收藏：8,000
  • 089.2亿
    symfony/polyfill-intl-normalizer
    本周热度排名：8 · Symfony polyfill for intl's Normalizer class and related functions · 累计下载：918,554,487 · 收藏：2,057
  • 097.8亿
    psr/http-factory
    本周热度排名：9 · PSR-17: Common interfaces for PSR-7 HTTP message factories · 累计下载：780,402,158 · 收藏：1,896
  • 1012.6亿
    psr/log
    本周热度排名：10 · Common interface for logging libraries · 累计下载：1,260,890,308 · 收藏：10,605
  • 119.0亿
    symfony/polyfill-php80
    本周热度排名：11 · Symfony polyfill backporting some PHP 8.0+ features to lower PHP versions · 累计下载：901,731,660 · 收藏：1,731
  • 128.1亿
    symfony/string
    本周热度排名：12 · Provides an object-oriented API to strings and deals with bytes, UTF-8 code points and grapheme clusters in a unified way · 累计下载：809,957,017 · 收藏：1,789
  • 1310.7亿
    symfony/polyfill-ctype
    本周热度排名：13 · Symfony polyfill for ctype functions · 累计下载：1,068,562,005 · 收藏：4,044
  • 1410.6亿
    guzzlehttp/promises
    本周热度排名：14 · Guzzle promises library · 累计下载：1,059,165,670 · 收藏：7,749
  • 1510.8亿
    guzzlehttp/guzzle
    本周热度排名：15 · Guzzle is a PHP HTTP client library · 累计下载：1,079,295,080 · 收藏：24,314
  • 167.9亿
    symfony/polyfill-intl-grapheme
    本周热度排名：16 · Symfony polyfill for intl's grapheme_* functions · 累计下载：786,576,646 · 收藏：1,725
  • 1710.6亿
    symfony/process
    本周热度排名：17 · Executes commands in sub-processes · 累计下载：1,058,525,794 · 收藏：7,513
  • 189.6亿
    ralouphie/getallheaders
    本周热度排名：18 · A polyfill for getallheaders. · 累计下载：961,240,334 · 收藏：3,799
  • 1910.9亿
    symfony/finder
    本周热度排名：19 · Finds files and directories via an intuitive fluent interface · 累计下载：1,085,955,352 · 收藏：8,512
  • 209.9亿
    nikic/php-parser
    本周热度排名：20 · A PHP parser written in PHP · 累计下载：986,067,096 · 收藏：17,526
  • 2110.7亿
    symfony/event-dispatcher
    本周热度排名：21 · Provides tools that allow your application components to communicate with each other by dispatching events and listening to them · 累计下载：1,069,519,633 · 收藏：8,577
  • 228.4亿
    symfony/event-dispatcher-contracts
    本周热度排名：22 · Generic abstractions related to dispatching event · 累计下载：837,940,521 · 收藏：3,415
  • 237.6亿
    psr/http-client
    本周热度排名：23 · Common interface for HTTP clients · 累计下载：762,634,298 · 收藏：1,732
  • 249.6亿
    symfony/http-foundation
    本周热度排名：24 · Defines an object-oriented layer for the HTTP specification · 累计下载：957,339,251 · 收藏：8,700
  • 2510.5亿
    monolog/monolog
    本周热度排名：25 · Sends your logs to files, sockets, inboxes, databases and various web services · 累计下载：1,045,039,950 · 收藏：22,263
  • 269.4亿
    symfony/var-dumper
    本周热度排名：26 · Provides mechanisms for walking through any arbitrary PHP variable · 累计下载：935,150,211 · 收藏：7,567
  • 277.0亿
    psr/event-dispatcher
    本周热度排名：27 · Standard interfaces for event handling. · 累计下载：696,209,700 · 收藏：2,290
  • 287.8亿
    symfony/translation-contracts
    本周热度排名：28 · Generic abstractions related to translation · 累计下载：776,985,179 · 收藏：2,586
  • 298.5亿
    symfony/polyfill-intl-idn
    本周热度排名：29 · Symfony polyfill for intl's idn_to_ascii and idn_to_utf8 functions · 累计下载：851,800,316 · 收藏：3,350
  • 307.5亿
    symfony/mime
    本周热度排名：30 · Allows manipulating MIME messages · 累计下载：745,330,742 · 收藏：2,833
  • 319.9亿
    doctrine/lexer
    本周热度排名：31 · PHP Doctrine Lexer parser library that can be used in Top-Down, Recursive Descent Parsers. · 累计下载：987,213,632 · 收藏：11,162
  • 329.1亿
    symfony/translation
    本周热度排名：32 · Provides tools to internationalize your application · 累计下载：911,599,855 · 收藏：6,639
  • 339.0亿
    symfony/http-kernel
    本周热度排名：33 · Provides a structured process for converting a Request into a Response · 累计下载：897,936,898 · 收藏：8,127
  • 348.0亿
    psr/simple-cache
    本周热度排名：34 · Common interfaces for simple caching · 累计下载：801,117,416 · 收藏：8,094
  • 354.1亿
    psr/clock
    本周热度排名：35 · Common interface for reading the clock. · 累计下载：406,617,831 · 收藏：646
  • 367.3亿
    symfony/error-handler
    本周热度排名：36 · Provides tools to manage errors and ease debugging PHP code · 累计下载：725,848,285 · 收藏：2,672
  • 379.2亿
    myclabs/deep-copy
    本周热度排名：37 · Create deep copies (clones) of your objects · 累计下载：921,501,169 · 收藏：8,910
  • 388.6亿
    symfony/routing
    本周热度排名：38 · Maps an HTTP request to a set of configuration variables · 累计下载：863,155,930 · 收藏：7,686
  • 397.7亿
    ramsey/uuid
    本周热度排名：39 · A PHP library for generating and working with universally unique identifiers (UUIDs). · 累计下载：772,832,253 · 收藏：12,854
  • 405.8亿
    brick/math
    本周热度排名：40 · Arbitrary-precision arithmetic library · 累计下载：575,722,158 · 收藏：2,178
  • 419.7亿
    sebastian/diff
    本周热度排名：41 · Diff implementation · 累计下载：973,207,917 · 收藏：7,701
  • 429.8亿
    phpunit/phpunit
    本周热度排名：42 · The PHP Unit Testing framework. · 累计下载：982,175,486 · 收藏：20,558
  • 439.3亿
    doctrine/inflector
    本周热度排名：43 · PHP Doctrine Inflector is a small library that can perform string manipulations with regard to upper/lowercase and singular/plural forms of words. · 累计下载：927,665,318 · 收藏：11,377
  • 449.5亿
    sebastian/exporter
    本周热度排名：44 · Provides the functionality to export PHP variables for visualization · 累计下载：952,718,591 · 收藏：6,820
  • 459.4亿
    sebastian/recursion-context
    本周热度排名：45 · Provides functionality to recursively process PHP variables · 累计下载：943,227,182 · 收藏：6,566
  • 467.6亿
    egulias/email-validator
    本周热度排名：46 · A library for validating emails against several RFCs · 累计下载：761,652,057 · 收藏：11,707
  • 479.5亿
    sebastian/comparator
    本周热度排名：47 · Provides the functionality to compare PHP values for equality · 累计下载：950,019,326 · 收藏：7,061
  • 489.6亿
    phpunit/php-code-coverage
    本周热度排名：48 · Library that provides collection, processing, and rendering functionality for PHP code coverage information. · 累计下载：962,399,487 · 收藏：8,970
  • 499.5亿
    phpunit/php-file-iterator
    本周热度排名：49 · FilterIterator implementation that filters files based on a list of suffixes. · 累计下载：948,900,389 · 收藏：7,487
  • 509.4亿
    sebastian/environment
    本周热度排名：50 · Provides functionality to handle HHVM/PHP environments · 累计下载：941,066,614 · 收藏：6,776
  • 519.3亿
    sebastian/global-state
    本周热度排名：51 · Snapshotting of global state · 累计下载：930,423,773 · 收藏：6,596
共 100 条热搜查看更多
远景论坛
Windows 11
1天前更新
  • 01
    用vivetool关闭新版右键菜单 系统26340.9212
    输入指令 .\\vivetool /disable /id:60813048 必须重启电脑生效47884125
  • 02
    26340启用包
    适合26300.9212转26340.9212，其他版本没测试！三人行1970
  • 03
    [8.19]29648.1000安装了一个，好像挺流畅的
    感觉比25H2的流畅啊。kernels
  • 04
    26340.9212 任务栏位置
    刚刚升级到 26340.9212，发现任务栏位置可以调了。 26300好像不能 的。win10_10139
  • 05
    暗夜黑白·DeCody 鼠标主题
    黑白碰撞，利落线条，操作自带速度感。告别花里胡哨，这套光标让你桌面气场全开。 自取↓ 🔗：https://pan.quark.cn/s/49e88b45cc4e 🔗：https://pan.xunlei.com/s/VP-NO7zPv3blqbJWP2R6vl2yA1?pwd=7qui# ...1757791488
  • 06
    Win11_26H2_26300.9267_IoT Enterprise LTSC 极度精简1.71G
    https://www.alipan.com/s/CCzBQ9GPMia 提取码: o6u5 MD5: D25F7768BA03714DDC85AF584595E39C SHA1: 069A20EC2560BDCB3C56A275143901297BBB0270 CRC32: 0041D5D8xiaofeng027
  • 07
    如果KMS不能用了，你会滑铲到Linux还是买正版
    gaozhe3321
  • 08
    自定义托盘图标顺序小工具
    可以不用把全部程序都打开就可以对托盘图标顺序进行排序，排好序之后要点击上方工具栏的保存再点击重启Explorer才能生效 右键对应行可以跳转到对应图标的目录或者删除这一个图标 下载链接：https://1827495994.share.123pan.cn/123pan/toj0jv-xanMh Github上也上传 ...修电脑大师
  • 09
    应用商店里面应用这个状态是什么意思？
    应用商店这些应用这个状态是什么意思？jackie-xu
  • 10
    [2026/8/19]英文 24H2/25H2/26H1 (2026年8月更新) x64&arm64 MVS ISO (zerofs网盘)下载
    arm64 名称：en-us_windows_11_consumer_editions_version_24h2_updated_aug_2026_arm64_dvd_c7a47ad8.iso SHA1校验： 5e9835544e476d8d8ec16d025c456883c5418271 下载地址：https://zerofs.link/f/HSiWsa7/ 名称：en-us_windows_11_consumer_editions_version_25h2_ ...jinggangshi
  • 11
    [2026/8/19]简体中文/繁体中文 26300_9212 官方iso 直链下载
    简体中文 繁体中文 连接有效期未知，失效不补 8e1809d2d505ae948a715c12d1dd6815c755161ea0cc0736d83351965bc526ec *Windows11_InsiderPreview_Client_x64_zh-cn_26300_9212.iso 5face90ce1a72b276ad41b245477812a4af5b07692118923c9fa50222dfe2541 *Windows11_Ins ...jinggangshi
  • 12
    [2026/8/19]英文 26300.9212 官方ISO 直链下载
    链接有效期未知，失效不补 Windows11_InsiderPreview_Client_x64_en-us_26300_9212.iso SHA1: 2c49dac095328628448c51c1a3598a5b4c6214f5 Windows11_InsiderPreview_EnterpriseVL_x64_en-us_26300_9212.iso SHA1: a6c7be60f903aaa71e60d0945d599e9bf44f63e6 ...jinggangshi
  • 13
    请问microsoft 照片如何升级到2026.11080.5006.0测试版？
    情况有点特殊： 家里台式机装的win11，26220.xxxx，自带的 microsoft 照片 版本是：2026.11080.5006.0 我笔记本装的win11，26200.xxxx，自带的 microsoft 照片 版本是：2026.11060.2004.0 发现台式机上的照片(2026.11080.5006.0)很好用，特别是信息的排版，简直顺眼太 ...conan304
  • 14
    壁纸
    壁纸，已经ai消除图标，并进行5k像素处理，4k屏放心食用。liwind
  • 15
    Win11 26xxx-29xxx，开启默认以最大化方式打开应用的选项
    开启默认以最大化方式打开应用的选项(26xxx-29xxx)： vivetool /enable /id:62915050rubycon
  • 16
    Windows 11 Insider Beta 预览版 26220.9202
    Windows 11 Insider Beta 预览版 26220.9202 ------------------------------------- 发布日期：2026年8月17日 2026年8月17日更新：已宣布从本测试版中移除Windows恢复增强功能，该功能已在实验版26340.9212中发布。 变更和改进 文件资源管理器 [*]现在启动文件资源管 ...小银狐
  • 17
    Windows 11 Insider Experimental Preview Build 26340.9212
    Windows 11 Insider Experimental Preview Build 26340.9212 ------------------------------------------ 发布日期：2026年8月17日 2026 年 8 月 17 日更新：将最初在 Beta 版本 26220.9202 中宣布的 Windows 恢复增强功能移至此实验版本。 各项变更和改进正在逐步推出 ...小银狐
  • 18
    Windows 11 Insider Beta (26H1) 预览版 28020.2731
    Windows 11 Insider Beta (26H1) 预览版 28020.2731 --------------------------- 发布日期：2026年8月17日 各项变更和改进正在逐步推出 [移除 Windows 管理规范命令行 (WMIC)] [*]此版本已移除 Windows 管理规范命令行工具 (WMIC)。此更改是 Windows 持续弃用和移除 W ...小银狐
  • 19
    Windows 11 Insider Experimental (26H1) 预览版 28120.2738
    Windows 11 Insider Experimental (26H1) 预览版 28120.2738 ------------------------- 发布日期：2026年8月17日 各项变更和改进正在逐步推出 [移除 Windows 管理规范命令行 (WMIC)] [*]此版本已移除 Windows 管理规范命令行工具 (WMIC)。此更改是 Windows 持续弃用和 ...小银狐
  • 20
    Windows 11 Insider Experimental（未来平台）预览版 29648.1000
    Windows 11 Insider Experimental（未来平台）预览版 29648.1000 ---------------------------------------------- 发布日期：2026年8月17日 各项变更和改进正在逐步推出 [Windows 安全] [*]我们正在对 Windows 安全中心应用进行一些小的用户界面调整，以使其与其他 Wi ...小银狐
共 20 条热搜查看完整榜单