近日，北京大学、清华大学、星尘智能等机构联合提出StableVLA，相关工作StableVLA: Towards Robust Vision-Language-Action Models without Extra Data已被ICML 2026接收 ...

PHNOM PENH, June 27 (Xinhua) -- Cambodia is committed to strengthening micro, small and medium enterprises (MSMEs) as a key pillar of economic growth, job creation, and national competitiveness, Minis ...

2026年2月，市科协发布《关于推荐“2026世界人工智能大会青年优秀论文奖”参评论文的通知》，面向全球高校、科研院所、企业的人工智能领域青年学者征集优秀论文。至征稿截止，共收到海内外参评论文195篇。经初评、复评、终评，专家评审委员会最终评选出2026世界人工智能大会青年优秀论文奖10篇，提名奖10篇。名单公示如下： 10. Structured 3D Latents for Scalable ...

IT之家6 月 30 日消息， When Jaiveer Singh talks about robots, he doesn’t begin with spectacle. He begins with infrastructure: the boards inside machines, the software that lets developers see through a robot ...

贾浩楠 发自 副驾寺智能车参考 | 公众号 AI4Auto理想，够燥的！前有力推的VLA路线被广泛质疑没用、走弯路，后有技术向的Big name们先后出走投身具身创业。但理想依然通过Livis Day向外界宣告：有人、有技术、有产品！

现代多模态AI（也就是能同时处理文字和图片的AI）本来就有一个接收图像的"视觉入口"。通常情况下，人们只是从这个入口送进去一张普通的照片，让AI描述或理解图片内容。但这个研究团队发现，如果送进去的不是普通照片，而是一张经过精心"设计"的特殊图片，是否能悄悄影响AI后续的文字回答？

GITEGA, June 24 (Xinhua) -- Africa-China relations exemplify model South-South cooperation, based on mutual respect, solidarity, and shared benefits, the current Chairperson of the African Union (AU) ...

Morning light filters through the carved wooden windows of a traditional Dai-ethnic stilt house, and golden flecks dance ...

NVIDIA Vera Rubin accelerates the world’s most demanding HPC workloads including climate modeling, computational fluid dynamics and energy exploration. NVIDIA Vera Rubin and Vera CPU to bring agentic ...

PJ TAYLOR talks with a highly qualified academic all-rounder who has a very friendly style and genuine interest in her ...

2026 年很值得关注的新 Benchmark ...