Professional graphic design is a long-horizon agentic task in which structured, editable a…
Occupational accident narratives contain valuable information about work situations, unfav…
Training capable coding agents via reinforcement learning (RL) requires diverse tasks with…
Memory systems for large language models have focused predominantly on efficient retrieval…
We introduce QuranicMMLU, a benchmark for evaluating generative AI on Quranic Arabic acros…
Vision-language models (VLMs) rely on storing and transferring appropriate information acr…
Computer-use agents (CUAs) have advanced along two separate lines: graphical interaction a…
Large language models can hold knowledge they do not report. A model may sandbag on a capa…
Electrocardiograms (ECGs) carry subject-specific patterns enabling reliable individual dis…
Long-term memory is essential for large language model (LLM) agents to maintain consistenc…
Interactive retrieval under partial evidence is a sequential information-acquisition probl…
Empirical evaluation is central to recommender-systems (RecSys) research, but turning expe…
Frontier language models now produce professional deliverables that expert graders judge t…
Multimodal survival models can combine complementary prognostic information from whole-sli…
We study how far a simple statistical pipeline can go on univariate time series anomaly de…
Emergency department (ED) decision-making relies on heterogeneous clinical information, in…
Behavioral failures can make a transformer appear to lack a world model even when it has l…
Conversational speech depends on dialogue context and the listener's immediately preceding…
Recent advances in large reasoning models (LRMs) have made machine unlearning more challen…
Large language models (LLMs) achieve strong performance but suffer from slow and costly in…
Large language models (LLMs) are increasingly deployed for enterprise information extracti…
On-policy distillation (OPD) improves reasoning models by learning the token-level discrep…
Access to academic support is a key determinant of student success, yet students experienc…
Fraudulent job posting detection aims to identify job advertisements that are corrupted ei…
Sequential recommendation aims to predict the next item a user will interact with based on…
Open-source native macOS AI reading workspace for PDF, EPUB, and TXT with highlights, note…
lint text in code by jev scorerer
Dotted, honestly-3D loading indicators for AI and agent interfaces in SwiftUI. Nine hand-t…
Hello Robot CEO and co-founder Aaron Edsinger will bring Stretch 4 for a live demo on the …
商汤U1 Pro，正式发布
The British AI data center developer depends on tech giants Microsoft and Anthropic for mo…
To build and deploy sophisticated robotics applications that can perceive, reason and act …
梦想是做能帮妈妈订机票的AI浏览器。
Hugging Face CEO：「太棒了」
日前，智能经济30人论坛AGI测评中心发布最新一期《AGI市场观察》（2026年9月12日-18日）。报告以OpenRouter词元消耗数据、各厂商官方披露价格及Arena基准测评…
9月22日，在2026杭州云栖大会企业级Agent实践峰会上，基元律动联合创始人兼CTO韩凯发表演讲《从Harness到RSI飞轮》。
上传成片和字幕，系统自动从字幕、音画、内容底线三个维度逐集跑一遍，输出可下载、可回填的质检报告。
9月22日，2026杭州云栖大会上，阿里云CTO李飞飞系统性阐述了Agentic Cloud 战略布局，阿里云将Model、Harness、Context作为核心构建场景，推出全新…
9月22日，在2026云栖大会「AI+文化传媒」技术发展论坛上，虎鲸文娱集团推出行业首个AI影视制作与管理平台“鲸锐AI”
9月22日，在2026年云栖大会上，千问宣布将加速打造 Personal Agent。千问产品负责人郑嗣寿表示，自去年上线以来，让每个人拥有自己的 AI 助理，始终是千问不变的目标…
<img data-src="https://static.leiphone.com/uploads/new/images/20260922/6ab234229a2d8.jpg" …
9 月 22 日，腾讯混元正式发布 Hy Image3.5 preview。腾讯视频专业影视智能平台 WorkRally 首发接入该模型，即日起两周内向平台创作者免费开放，覆盖剧集…
9月22日，新一代全系列千问 AI 硬件在云栖大会首次集中亮相，包含千问AI眼镜N1、N1 Pro及千问AI耳夹式耳机。三款新品当天同步开启线上预约，并将于10月13日现货发售。 …
未来Qwen4.5、Qwen5等版本将扩展至5-10T参数。
9月22日， 2026云栖大会开幕，阿里巴巴公布大模型最新进展。
为进一步深化人工智能在全球健康领域的创新实践，共同应对全球人工智能健康应用面临的挑战，人工智能促进健康全球倡议2026年会于9月16-18日在浙江杭州举办。本次会议由世界卫生组织（…
外界眼里，FDE火得一塌糊涂；一线从业者看到的， 却是荒诞、滑稽和不可思议。 前段时间，To B老将王涛看到某家号称在做FDE的厂商，公司内部一天的Token消耗量也就100多亿。…
吴泳铭表示，阿里巴巴将坚定投入AI模型、AI芯片和AI云这三项基础设施建设，这是阿里的长期战略选择。
9月22日，在2026杭州云栖大会上，阿里巴巴集团CEO吴泳铭发表最新演讲，阐述了对“机器智能”时代的思考。
医疗影像设备长期存在一个矛盾：算力需求持续上涨，机箱空间却难以增加。 PCIe 显卡越来越强，图像采集卡也不可或缺，但小型化 Mini-ITX 主板通常只提供一个 PCIe 插槽，…
弥补算力缺口，不能只靠堆料
9月17日，一家头部手机厂商亮相一款9.8克的球形AI记录设备，可夹衣佩戴，宣称续航24小时。该产品一经曝光便在全网疯狂刷屏，“手机能录音，为什么还需要独立硬件”，再次成为行业热议…
以超声写脑、多模态读脑、AI神经解码构建闭环，建设中国非侵入AI脑机接口标杆企业 <p style="text-align: justify; margin-bottom: 8p
The group won't be given leeway to slow down or redirect OpenAI's ongoing mathematical res…
Meta’s new AI agent Muse has racked up more downloads and daily active users in the U.S. a…
Every AI factory needs power and cooling that fit its computing architecture. As AI infras…
Muse isn't welcome as a shopper anymore.
Tabby is designed to be a real-time bookkeeping interface, handling clients’ paperwork as …
Physical AI is moving rapidly from research to large-scale deployment. By 2035, ABI Resear…
Today, Egypt’s AI builders gathered in the Grand Egyptian Museum for a reception that high…
OpenAI is working with an independent Advisory Group on Mathematics and Artificial Intelli…
With GPT-6 Astra, Higgsfield AI makes video ad creation easier for small businesses and br…
OpenAI outlines a path to shared global AI standards, calling for coordinated evaluation, …
Explore new OpenAI Academy learning paths for employees, developers, leaders, educators, a…
YouTube · Two Minute Papers
YouTube · Google for Developers
YouTube · OpenAI
YouTube · OpenAI
YouTube · OpenAI
YouTube · OpenAI
YouTube · OpenAI