[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"tag-vector-reward":3},{"tag":4,"articles":10},{"id":5,"name":6,"slug":7,"article_count":8,"description_zh":9,"description_en":9},"1d78b949-3248-4635-9147-6e16bd09ba9d","vector reward","vector-reward",0,null,[11],{"id":12,"slug":13,"title":14,"summary":15,"category":16,"image_url":17,"cover_image":17,"language":18,"created_at":19},"bc5e34d0-1e65-4ce0-a130-424da873bd5a","vector-policy-optimization-search-diversity-zh","VPO 讓搜尋更會找答案","VPO 把語言模型訓練成會產生多樣解法，讓 test-time search 在多候選搜尋時更容易挑到好答案。","research","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1779432361030-ief3.png","zh","2026-05-22T06:45:29.893413+00:00"]