实时 AI 消息
按发布时间追踪大模型、AI Agent、开源项目、算力基础设施、政策和产品更新,并沉淀到每日归档。
AI动态:Apple’s Ternus era begins as Nvidia bets on the whole AI stack
据 techcrunch.com 消息,Apple’s Ternus era begins as Nvidia bets on the whole AI stack。It’s officially the Ternus era at Apple. Tim Cook stepped down as CEO this week, handing the company to former hardware chief John Ternus, whose first memo promised a “huge launch next week” — timing that puts Apple’s next iPhone event on his desk before he’s...

AI动态:Meta is paying to peek at how you use their latest AI model
据 techcrunch.com 消息,Meta is paying to peek at how you use their latest AI model。For its new Muse Spark model, intended for operating coding and other agents, Meta is offering an explicit discount averaging out to about 95% for users who "contribute" to the development of future models by sharing their prompts and model outputs.

英伟达在 IFA 2026 加速本地 AI:RTX Spark Windows PC 十月上市,PAIR 工具可串联多台 PC 算力
在 IFA 2026 上,英伟达联合微软等伙伴集中发布了本地 AI 升级:Hermes Agent、OpenClaw、Perplexity Portable Computer 将获得简化的本地模型配置,llama.cpp 推理吞吐最高提升 1.9 倍,还推出了可在局域网内调度多台 PC 算力的免费工具 NVIDIA PAIR。联想与宏碁的紧凑型 NVIDIA RTX Spark Windows PC 将于 10 月上市。

Hcompany 开源 NeoMME:单个双向 Transformer 处理图文的多语言多模态编码器
9 月 3 日,Hcompany 团队开源 NeoMME:260M 与 800M 两种规模的多语言多模态编码器,用一个双向 Transformer 同时处理文本与图像,无需独立视觉塔或因果语言模型,权重以 Apache 2.0 发布并原生接入 Transformers。其检索版 NeoMME-Retriever 在 ViDoRe v3 以 260M 参数取得 0.523 nDCG@10,并把每页晚交互索引从约 1.5 MB 压缩到 6 kB,仍保留 95% 以上检索质量。

Daily Archive












