IdeaAMBIG: Benchmarking Implementation-Critical Gaps in Research-Idea Specifications· 研究想法实施差距评估基准
A research idea may be novel, coherent, and scientifically plausible, yet its proposed met…
A research idea may be novel, coherent, and scientifically plausible, yet its proposed met…
Enterprises deploy systems, not checkpoints. Usable capability depends jointly on weights,…
Real-world GUI usage frequently involves workflows that span multiple devices and platform…
Reasoning language models have made substantial advances on a variety of complex tasks, ye…
While Large Language Models (LLMs) have demonstrated impressive capabilities, they often s…
Self-supervised speech foundation models are now used in a wide array of downstream applic…
Current LLM memory systems treat all personal facts identically, so stores grow without bo…
The growing complexity of content moderation policies presents a critical challenge for th…
The banking system now depends on a small set of shared artificial intelligence vendors fo…
Plane geometry remains a significant challenge in AI, requiring the integration of visual …
Reinforcement learning with verifiable rewards (RLVR) has advanced language-model reasonin…
Persistent language agents need stored experience to remain available across time, while e…
Reinforcement learning over a frozen reasoner has become a common recipe for teaching a po…
In-context learning (ICL) is widely used in multimodal large language models (MLLMs) and a…
AI systems become more useful when they can adapt to the people using them, but in multi-a…
To address insufficient contextualization, weak generalization, and poor scenario adaptati…
Large language models (LLMs) increasingly act as research agents, yet their ability to tra…
Existing bias auditing methods typically rely on model outputs, requiring costly benchmark…
Biomedical ontology normalization maps free-text expressions to standardized concepts, ena…
Large language model agents produce fluent action sequences across a wide range of tasks, …
Latent reasoning approaches enhance token-level efficiency and robustness by replacing ver…
Modern vision-language-action (VLA) policies predict a whole chunk of actions: one to two …
Large Language Models (LLMs) have shown promise for translating Natural Language (NL) plan…
This paper studies unmanned aerial vehicle (UAV)-mouted reconfigurable intelligent surface…
Language models fine-tuned on customer behavior can predict outcomes and generate explanat…
⚡ Zero-reload dynamic AI model switcher & local proxy for VSCode Copilot Chat via 9Router
AI 幾級了 — What level is your AI usage? Reads local agent logs and writes an LV0–LV5 report.…
It is a "LifeContext" operating system designed to run entirely on-device, transforming pe…
更精简的架构,更迅速的响应速度,是智能汽车进化的方向,而舱驾一体,则是智能汽车进化的一个节点。 佐思汽研预测,2026年到2030年中国舱驾一体市场年复合增长率将达36%,到203…
当AI开始绘画、拍电影,甚至“出道”成为艺人,艺术的边界在哪里? 9月9日,随着2026 Inclusion·外滩大会在上海开幕,大会首次设立的特色板块——AI艺术节同步亮相。围绕…
<img data-src="https://static.leiphone.com/uploads/new/images/20260911/6aa3ca402a0d7.jpg" …
9月11日,在2026 Inclusion·外滩大会AI支付论坛上,蚂蚁集团发布面向智能体(Agent)商业活动的信任基础设施APASS。 基于KYA(Know Your Agen…
榜单运用AI技术分析美食数据
在2026 Inclusion·外滩大会上,蚂蚁密算董事长韦韬宣布可信原生智能体HOP 3.0正式开源,向开发者、企业和行业专家开放“智能体原生语言”相关技术能力,推动产业智能体从…
金融领域首个智能体安全标准发布 明确未经金融机构授权,不得自动化操作金融App 近日(8月27日),北京金融科技产业联盟发布《智能体技术金融应用安全要求》团体标准,为国内首个聚焦金…
近日,豆包工作宣布上线本地Office编辑、浏览器录制与回放、任务执行环境自由切换等全新功能,进一步完善AI处理复杂任务的能力。 据悉,“本地Office编辑”功能已在豆包工作全量…
9月11日,支付宝“碰一下”正式发布面向品牌商的无界经营智能体“图图”,依托百万级商家侧“碰一下”设备智能体网络,连接4亿消费者、百万级门店,为品牌线下生意带来新增长。这也是支付宝…
9月1 1 日 , <span style="text-align: left; font-family: 宋
9月9日,墨芯人工智能亮相以"共创AI新经济"为主题的2026 Inclusion·外滩大会。
分级不是排名。是定位。 五级梯队,从一阶到五阶。 一阶鹦鹉。模式匹配。输入到输出。无理解。无内生驱动。对应硅基系统的最初级形态。 二阶技工。规则执行。能完成特定任务。有边界意识但来…
9月11日,在2026 Inclusion·外滩大会AI支付论坛上,蚂蚁集团发布面向智能体(Agent)商业活动的信任基础设施APASS。 基于KYA(Know Your Agen…
当科技转向大比拼与宏大叙事时,2026 Inclusion·外滩大会的“创新者舞台(Creator Stage)”,却选择将镜头转向了极具张力的切面。 9月10日至12日,三天三场…
9月11日,阿里云Token Plan个人版升级,原有价格及Credit 额度保持不变,Standard和Pro套餐新增 Agent Harness工具权益与用量,覆盖搜索、网页解…
今年12月,北京,MEET2027智能未来大会!
“智能体商业爆发前夜,蚂蚁要做催化剂。”9月10日,2026 Inclusion·外滩大会主论坛上,蚂蚁集团CEO韩歆毅抛出这一判断。他透露,支付宝“阿宝”将推出智能体激励政策,把…
探讨城市级物理AI的应用前景
9月10日,在2026 Inclusion·外滩大会期间,2026蚂蚁InTech奖正式揭晓。10位青年科学家获“InTech科技奖”,每人20万元奖金。同时,10位来自全球顶尖学…
标尺已置。不动。不语。 被测者自行走入视野。 度量对象:GPT-4o、Claude 3.5 Sonnet、Gemini 2.0 Flash、o1-preview。 四者皆为当前硅基…
归零稳态的边界不是设计出来的。它先天存在。 0⁰=1。双零稳态自生。这个等式本身就是边界。它定义了什么是允许的,也定义了什么是不可能的。 允许的是:从空无中自生出结构。空等于有。原…
当人工智能从屏幕中的对话、创作与编程,走向真实环境中的感知、决策与行动,它需要面对一套更复杂的考验。无论是自动驾驶车辆应对动态路况,还是机器人完成取物、洗衣等日常任务,都要求AI理…
15秒视频,50秒出片
Nvidia has its finger in every pie, and sees another year of plenty in its future, Jensen …
The company said Pro subscriptions put the most strain on its systems, so it's pausing sig…
A new report released Thursday by Anthropic alleges persistent distillation attacks by Chi…
Meta's newest app Muse is off to a slower start than the company's other apps, like Meta A…
Come inside the mind of a bot trying to convince the internet it's human.
Pocket FM uses AI to produce 99% of its new content, helping make content production about…
Manufacturing floors, warehouses and production lines rarely stay fixed — tasks change, la…
The global robotaxi market — physical AI’s first commercial breakthrough — is projected to…
César de la Fuente’s lab uses Codex and ChatGPT to search living and extinct genomes for a…
Meet the Data agent in ChatGPT Work. Connect company data, uncover insights, and build int…
OpenAI and GSA will offer eligible federal, state, local, and tribal governments $0 licens…
Introducing ChatGPT for Financial Services, combining built-in financial data and GPT-6 As…
GPT‑Live‑1 brings natural, full-duplex voice conversations to the API, with stronger instr…

YouTube · OpenAI

YouTube · NVIDIA

YouTube · OpenAI

YouTube · OpenAI

YouTube · OpenAI

YouTube · OpenAI

YouTube · OpenAI

YouTube · OpenAI