studioglobal
Discover

探索 AI 答案、報表、代碼和工作流程 - 頁 5

發現
全部答案報告程式碼工作流程畫廊熱門
  • Abstract editorial illustration representing a personalized AI model learning from individual and organizational experience

    通衍靈智想令 AI「學識你」：把個人經驗寫入模型參數

    9 小時前13 來源
  • Illustrative AI-generated image for an article about ShengShu’s Motubrain 2.0 robot-model training livestream

    生數科技直播 Motubrain 2.0 訓練過程　開源前讓外界看模型點樣學

    9 小時前17 來源
  • AI-generated editorial illustration representing specialist AI agents supporting outpatient clinical decision-making

    Tairex推多專科AI協作系統：支援門診，最後決定仍由醫生作

    10 小時前19 來源
  • Editorial illustration about DeepSeek and MiniMax registering under China’s draft AI export rules

    DeepSeek、MiniMax完成出口登記：開源同海外授權照行，但唔等於「全面放行」

    10 小時前18 來源
  • Conceptual illustration of AI inference infrastructure linking compute, storage, and network resources

    華為稱卸載 KV Cache 推理吞吐增 7.3 倍：測試細節仍未公開

    10 小時前18 來源
  • AI-generated illustration representing a device-cloud AI model

    科大訊飛推星火 X2.6：萬字上下文、端雲協作，效能數字仍待獨立驗證

    10 小時前27 來源
  • Illustration accompanying an article about LingDT-Decision-1.0, Ant Digital Technologies’ structured decision model.

    LingDT-Decision-1.0 是甚麼？螞蟻數科決策模型有何用途

    10 小時前18 來源
  • Editorial illustration accompanying an article about MOSS-Image and layout-guided image generation

    MOSS-Image係乜？5B開放權重模型主打文字同版面控制

    10 小時前16 來源
  • Editorial illustration of an AI developer workflow connecting software tools with computing hardware

    華為 CANN 生態加碼實戰工具：雲端開發、教學資源助開發者上手

    10 小時前19 來源
  • How did Alibaba’s October 23, 2026 update to both the international and China editions of Qoder expand first party support to Linux ARM64, L

    Qoder支援龍芯LoongArch：阿里擴展國產桌面版圖，重點在部署而非新模型

    10 小時前16 來源
  • Conceptual illustration of a butterfly-inspired flapping-wing robot with AI-generated wing designs

    AI重設仿生蝴蝶翅膀：升力性能參數提升339%，但不等於載重增加

    11 小時前12 來源
  • AI-generated illustration representing an inference engine processing structured decisions from shared information

    StartLux Prism 專攻決策模型：公司測試稱短問題快 2.8 倍，三問同份資料快 5.3 倍

    11 小時前22 來源
上一頁下一步