[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"tag-google-research":3},{"tag":4,"articles":11},{"id":5,"name":6,"slug":7,"article_count":8,"description_zh":9,"description_en":10},"096cc99c-29af-47c1-b451-7d7f10424d2d","Google Research","google-research",4,"Google Research 聚焦模型推論、提示設計與系統效率等基礎問題，從記憶體壓縮到輸入策略都會影響成本與準確率。這類研究常直接改變 AI 服務的部署方式與評估方法。","Google Research covers the core engineering behind modern AI: inference efficiency, memory compression, prompt behavior, and benchmark reliability. Its work often shapes how models are deployed, measured, and optimized in production.",[12,21],{"id":13,"slug":14,"title":15,"summary":16,"category":17,"image_url":18,"cover_image":18,"language":19,"created_at":20},"82766fdc-4368-445d-bb4a-03377726df02","turboquant-cuts-memory-use-without-accuracy-loss-zh","TurboQuant 省 6 倍記憶體，還不掉準確率","Google Research 發表 TurboQuant，主打記憶體用量降到 1\u002F6、推論快 8 倍，且在報告測試中沒有準確率損失。這篇看它怎麼改 AI 伺服器成本。","research","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1775161134112-ftrj.png","zh","2026-04-02T20:18:39.266389+00:00",{"id":22,"slug":23,"title":24,"summary":25,"category":17,"image_url":26,"cover_image":26,"language":19,"created_at":27},"87335969-ee48-4021-bd27-6731750537ff","duplicate-prompts-can-lift-accuracy-fast-zh","重複提示詞，準確率真的會上升","Google Research 研究發現，提示詞複製一次可讓 70 組模型與基準測試中的 47 組提升準確率，NameIndex 甚至從 21.33% 衝到 97.33%。","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1775122500397-vvmh.png","2026-04-02T08:39:34.363421+00:00"]