Yuyao Zhang, Junjie Gao, Zhengxian Wu, Jiaming Fan, Jin Zhang, Shihan Ma, Yao Yao, Weiran Qi, Chuyan Jin, Guiyu Ma, Xingzhong Xu, Kai Yang, Ji-Rong Wen, Zhicheng Dou · 2026-07-18 · 3 min AI

SearchOS-V1: 实现强大的开放域信息搜索代理协作

SearchOS-V1: Towards Robust Open-Domain Information-Seeking Agent Collaboration

工具集成大型语言模型的最新进展使网络搜索成为信息查找代理的核心能力。然而,作为交互...

Recent advances in Tool-Integrated Large Language Models have made web search a core capability of information-seeking agents. However, as interaction...

01
Victoria Graf, Hannaneh Hajishirzi, Noah A. Smith, David Kohlbrenner, Kyle Lo · 2026-07-18 · 9 min AI

预训练数据可能会通过计算宣传而被毒害

Pretraining Data Can Be Poisoned through Computational Propaganda

破坏预训练数据可能会给 LM 带来难以检测和缓解的有害行为。先前关于中毒预训练数据的工作......

Poisoning pretraining data can introduce harmful behaviors to LMs that are difficult to detect and mitigate. Prior work on poisoning pretraining data ...

02
Michaela Jarvis | MIT Laboratory for Information and Decision Systems · 2026-07-18 · 10 min AI

追随问题的引导

Following the questions where they lead

贝利·弗拉尼根(Bailey Flanigan)从小就在威斯康星州,自家’的农田里玩耍,她的选择性,而又广泛的,好奇心引导着她……

Ever since she was a child playing on her family’s farmland in Wisconsin, Bailey Flanigan was guided by her own selective, yet wide-ranging, curiosity...

03
Adam Zewe | MIT News · 2026-07-17 · 9 min AI

将 2D 设计转变为 3D 模型以进行快速原型设计的更好方法

A better way to turn 2D designs into 3D models for rapid prototyping

工程师经常使用视觉语言模型来产生新的设计,,例如飞机或汽车部件的设计。为了模拟这些组件将如何...

Engineers often use vision-language models to produce new designs, such as for airplane or automobile components. To simulate how those components wil...

04
Harsha Vardhan Khurdula, Abhinav Kumar Singh, Yoeven D Khemlani, Vineet Agarwal · 2026-07-16 · 4 min AI

使用冻结的离散扩散语言模型进行音频本机语音识别

Audio-Native Speech Recognition with a Frozen Discrete-Diffusion Language Model

自动语音识别主要由一次发出一个标记的自回归解码器主导。我们问离散扩散语言模型是否......

Automatic speech recognition is dominated by autoregressive decoders that emit one token at a time. We ask whether a discrete diffusion language model...

05
Junjie Yin, Xinyu Feng · 2026-07-16 · 4 min AI

AI 代理是否知道任务何时简单? 进行复杂性感知推理和执行

Do AI Agents Know When a Task Is Simple? Toward Complexity-Aware Reasoning and Execution

大型语言模型 (LLM) 代理越来越多地实现多步骤工程和信息学工作流程的自动化,,但他们很少询问一项任务需要付出多少努力...

Large language model (LLM) agents increasingly automate multi-step engineering and informatics workflows, yet they rarely ask how much effort a task a...

06
Media Lab · 2026-07-16 · 10 min AI

3 个问题: 神经透明度和 AI 设计的未来

3 Questions: Neural transparency and the future of AI design

现在,数百万人正在设计自己的个性化人工智能伴侣,,但大多数人几乎不知道这些创作将如何发挥作用......

Millions of people are now designing their own personalized artificial intelligence companions, yet most have little idea how those creations will act...

07
Department of Aeronautics and Astronautics · 2026-07-15 · 6 min AI

AI 能否建造喷气发动机? JARVIS 挑战赛测试 AI 副驾驶在艰难环境中的作用

Can AI build a jet engine? JARVIS Challenge tests role of AI copilots in tough

人工智能迅速改变了软件工程。生成式 AI 和大型语言模型 (LLMs) 可以创建大量代码...

Artificial intelligence has rapidly transformed software engineering. Generative AI and large language models (LLMs) can create huge volumes of code a...

08
David Chandler | Laboratory for Information and Decision Systems · 2026-07-15 · 3 min AI

帮助AI模型满足现实世界

Helping AI models to meet the real world

近年来,使用人工智能来增强预测, 规划, 和企业决策的系统一直在激增, 但...

Systems using artificial intelligence to enhance forecasting, planning, and decision-making in businesses have been proliferating in recent years, but...

09
yxc0433 · 2026-07-14 · 4 min US-China Trade

CNBC的 《中国连线》时事通讯: 押注深圳而非硅谷

CNBC's The China Connection newsletter: Betting on Shenzhen over Silicon Valley

星期一 - 星期五, 10:00 - 11:00 SIN/HK | 0400 - 05:00 CET 你好,,这是伊芙琳, 从北京写给您的信。欢迎阅读最新一期的中国...

Monday - Friday, 10:00 - 11:00 SIN/HK | 0400 - 05:00 CET Hi, this is Evelyn, writing to you from Beijing. Welcome to the latest edition of The China C...

10