Surprisal Theory is Tautological (without Rational Grounding)· 惊奇理论是同义反复(无理性依据)
Surprisal theory holds that the human processing difficulty of a linguistic unit in contex…
Surprisal theory holds that the human processing difficulty of a linguistic unit in contex…
Large Language Models (LLMs) show promise for medical education, but most existing systems…
A consensus anomaly detection framework was applied to monthly malaria surveillance data f…
Building socially calibrated large language models, which can learn from others without si…
Modern AI agents rely on elaborate inference harnesses such as Claude Code, Codex, and Ope…
Unlike large language models (LLMs) that exhibit strong reasoning capabilities, vision-lan…
The rapid progress of AI has intensified the long-standing pursuit of automation: replacin…
We present DONDO, a family of open, permissively licensed automatic speech recognition (AS…
Even a current high-capability LLM can appear safer when shown a dangerous objective direc…
Production AI agents' failures are less often due to an inability to reason well and more …
A rhetorical figure that Cicero and Quintilian catalogued two thousand years ago reappears…
AI agents are increasingly created inside organizations by non-engineering users through l…
Do independently trained language models come to represent the same thing in the same way?…
Large language models (LLMs) and agents are now widely used tools in code development, wit…
Deep research requires agents to find answers that jointly satisfy multiple constraints. D…
The rise of human-AI collaborative writing has created a growing need for fine-grained det…
Deep learning models can effectively use Rapid Evaporative Ionization Mass Spectrometry (R…
Generative models can support decision-making under uncertainty by producing ensembles of …
In long-horizon LLM agent reinforcement learning, weak policies often repeat similar failu…
In automated planning, logical regression is an operation that returns the most general co…
Large Language Models (LLMs) excel at natural language understanding and generation but re…
Self-supervised foundation models have recently shown strong potential for electroencephal…
Electroencephalography (EEG) models used for epilepsy are often limited to specific datase…
Instance-level explanations aim to reveal the rationale behind a model's decisions for a s…
For years, supply chain planning at e-commerce firms has operated as a collection of isola…
A curated collection of offensive, defensive and AI/LLM security tools.
Flatkey media generation CLI for images, videos, audio, text, credits, and model discovery…
Run the pi coding agent as a service — triggered on demand, on a cron schedule, or by a Gi…
开源、隐私、本地优先、模型无关
这是机器人领域最好的时代
不做人形,抓住真实用户需求
工业圈和学术圈“唱反调”
文|温丽虹 编辑|张雨忻 Kimi K3上线前夕,曾歆勋算了算,离职时放弃的那笔期权的价值,在他离开这段时间涨了10倍。 从月之暗面离职前,曾歆勋是Kimi的AI搜索技术负责人。2…
英伟达宣布,SK集团与英伟达在AI工厂及下一代内存领域扩大战略合作。SK集团与英伟达今日宣布计划建立一项超过5000亿美元的综合合作伙伴关系,以建立满足全球计算需求激增的人工智能基…
深圳,7月23日——Hi3D实体创意生态闭门会在深圳举行。围绕“探索AI驱动的下一代可制造生态”,Hi3D与3D打印硬件、制造、创作者及Maker领域嘉宾共同探讨AI 3D从数字生…
7月24日,TCL科技发布公告,公司收购广州华星半导体45%股权事项已获深交所审核通过。根据本次交易方案,收购完成后TCL科技将直接及间接持有广州华星半导体100%股权。据了解,广…
单卡突破10000 video tokens/s,Vivix打通实时多模态生成全链路
随着全球科技巨头竞相加码人工智能基础设施建设,穆迪评级警告称,AI投资热潮正侵蚀大型云计算服务商的自由现金流,并推高资产负债表风险,未来投资者将更加关注这些企业能否从巨额AI投入中…
Wind数据显示,截至7月24日记者发稿时,6月以来已有19家上市券商及券商母公司累计接待了43次机构调研。与以往聚焦经纪和资管业务不同,如今调研机构最关心的是各家券商在科创领域的…
OpenAI:我们注意到,围绕抱抱脸系统(Hugging Face)被入侵事件流传着许多疑问和各种猜测。我们正与外部顾问合作,并在安全与安保委员会的监督下,继续进行全面审查。审查完…
当地时间7月24日,Anthropic推出新模型Opus 5。公司称,Opus 5的性能接近其更强大的同系列产品Fable 5,而价格却只有后者的一半,适用于日常办公和计算机编程任…
据报道,英伟达表示,将协助海力士设计未来的高带宽内存芯片,此举将有助于确保内存供应。英伟达还表示,与SK集团合作的“人工智能计划”总价值将超过5000亿美元。黄仁勋在采访中表示,这…
今日热点导览 小鹏主动召回3.35万辆X9,涉及前空气弹簧安全隐患 微信撤回消息后可删除提示字 怡颗莓莓果线下仍在售,公司称针对近期舆论正准备维权 菲尔兹奖得主雅各布·齐默曼将加入…
OpenAI's fancy new AI keypad will be a lot of fun for some, while many others are probably…
The neolab is betting that automating routine computer tasks will soon outpace coding as A…
Opus 5 在人工分析智能排行榜上位居第一。
AI coding startup Cognition has acquired Poke, the AI assistant you text like a friend, in…
Opus 5 will be both cheaper and less restrictive than Fable, likely making it preferable i…
AI companies including Nvidia and Mistral urge policymakers to avoid broad restrictions on…
Users can now ask Attie questions about news, trends, and conversations on Bluesky and oth…
The AI lab Midjourney continues to expand its purview beyond image and video generation.
<img class="rich_pages wxw-img" data-aistatus="1" data-croporisrc="https://mmbiz.qpic.cn/m…
跑出原生全模态新路径
给它一张图,还你整个世界
如今线上题库、网课、自适应习题工具早已普及,但教育行业质量、成本、规模难以兼顾的“不可能三角”依然尚未打破。 同一班级里,有的学生一点就通、有的卡在基础概念、有的陷入思维误区,统一…
Chinese AI lab Moonshot’s open model Kimi went viral this week for reasons that had less t…
扎克伯格新目标:研发通用人工智能
解释为何通用 AI 难以实现
DeepMind与Google在AI领域的控制权争夺。
算法定价与AI共谋行为探讨。
芬兰提供免费在线 AI 课程,面向全球开放。
Mark Cuban 强调学习 AI 的重要性
顶级扑克选手对决AI
AI 编程挑战赛,提升算法能力
乔姆斯基谈AI发展误入歧途

YouTube · NVIDIA

YouTube · NVIDIA

YouTube · OpenAI

YouTube · NVIDIA