[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"tag-nlg-evaluation":3},{"tag":4,"articles":10},{"id":5,"name":6,"slug":7,"article_count":8,"description_zh":9,"description_en":9},"c8b7e221-bbdd-40a8-a7ed-351ba2e06557","NLG evaluation","nlg-evaluation",2,null,[11],{"id":12,"slug":13,"title":14,"summary":15,"category":16,"image_url":17,"cover_image":17,"language":18,"created_at":19},"3a330546-beae-4173-9b71-9d0d446ff432","llm-judge-reliability-conformal-transitivity-en","How to Trust LLM Judges, Per Input","A diagnostic toolkit shows LLM judges can look stable on average while still being unreliable on individual inputs.","research","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1776406194842-ml7e.png","en","2026-04-17T06:09:33.415041+00:00"]