{"id":6883,"date":"2025-06-09T18:00:00","date_gmt":"2025-06-09T09:00:00","guid":{"rendered":"https:\/\/devneko.jp\/wordpress\/?p=6883"},"modified":"2025-06-07T21:05:18","modified_gmt":"2025-06-07T12:05:18","slug":"quantitative-llm-judges","status":"publish","type":"post","link":"https:\/\/devneko.jp\/wordpress\/?p=6883","title":{"rendered":"Quantitative LLM Judges"},"content":{"rendered":"\n<ul class=\"wp-block-list\">\n<li><strong>Quantitative LLM Judges\u00a0<\/strong>[48.7]<br>\u672c\u7814\u7a76\u3067\u306f,\u65e2\u5b58\u306eLLM\u5be9\u67fb\u54e1\u306e\u8a55\u4fa1\u30b9\u30b3\u30a2\u3092,\u4e0e\u3048\u3089\u308c\u305f\u9818\u57df\u306b\u304a\u3051\u308b\u4eba\u9593\u306e\u8a55\u4fa1\u30b9\u30b3\u30a2\u3068\u6574\u5408\u3055\u305b\u308b\u5b9a\u91cf\u7684LLM\u5224\u5b9a\u8005\u3092\u63d0\u6848\u3059\u308b\u3002 \u30e2\u30c7\u30eb\u306f\u3001\u88c1\u5224\u5b98\u306e\u30c6\u30ad\u30b9\u30c8\u8a55\u4fa1\u3068\u30b9\u30b3\u30a2\u3092\u7528\u3044\u3066\u3001\u539f\u5224\u4e8b\u306e\u30b9\u30b3\u30a2\u3092\u6539\u5584\u3059\u308b\u305f\u3081\u306b\u8a13\u7df4\u3055\u308c\u308b\u3002 \u5b9f\u9a13\u306b\u3088\u308a, \u5b9a\u91cf\u7684\u306a\u5224\u65ad\u306f, \u30dd\u30b9\u30c8\u30db\u30c3\u30af\u30e2\u30c7\u30ea\u30f3\u30b0\u306b\u3088\u308a, \u65e2\u5b58\u306e\u5224\u65ad\u306e\u4e88\u6e2c\u529b\u3092\u52b9\u679c\u7684\u306b\u5411\u4e0a\u3067\u304d\u308b\u3053\u3068\u304c\u793a\u3055\u308c\u305f\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2506.02945v1\">\u8ad6\u6587<\/a>\u00a0\u00a0<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2506.02945v1\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>\u00a0 \u00a0(Tue, 03 Jun 2025 14:44:23 GMT)<\/li>\n\n\n\n<li>\u300cWe introduce quantitative judges, a family of LLM judges that disentangle qualitative reasoning from quantitative score prediction in LLM-as-a-judge. Our approach has two stages: the qualitative stage, where a frozen LLM judge generates an evaluation, and the quantitative stage, where these outputs are used by a lightweight model to predict a human score.\u300d\u3068\u3044\u3046\u30a2\u30d7\u30ed\u30fc\u30c1\u306b\u3088\u308b\u5b9a\u91cf\u8a55\u4fa1<\/li>\n\n\n\n<li>\u73fe\u5b9f\u7684\u306a\u8a2d\u8a08\u65b9\u91dd\u306b\u601d\u3048\u308b\u3002<\/li>\n<\/ul>\n","protected":false},"excerpt":{"rendered":"","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[2],"tags":[224],"class_list":["post-6883","post","type-post","status-publish","format-standard","hentry","category-arxiv","tag-llm-as-a-judge"],"_links":{"self":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/6883","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=6883"}],"version-history":[{"count":1,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/6883\/revisions"}],"predecessor-version":[{"id":6884,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/6883\/revisions\/6884"}],"wp:attachment":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=6883"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=6883"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=6883"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}