{"id":5549,"date":"2024-10-07T05:07:00","date_gmt":"2024-10-06T20:07:00","guid":{"rendered":"https:\/\/devneko.jp\/wordpress\/?p=5549"},"modified":"2024-10-07T05:07:00","modified_gmt":"2024-10-06T20:07:00","slug":"justice-or-prejudice-quantifying-biases-in-llm-as-a-judge","status":"publish","type":"post","link":"https:\/\/devneko.jp\/wordpress\/?p=5549","title":{"rendered":"Justice or Prejudice? Quantifying Biases in LLM-as-a-Judge"},"content":{"rendered":"\n<ul class=\"wp-block-list\">\n<li><strong>Justice or Prejudice? Quantifying Biases in LLM-as-a-Judge\u00a0<\/strong>[84.3]<br>\u591a\u304f\u306e\u9818\u57df\u3067\u512a\u308c\u3066\u3044\u308b\u306b\u3082\u304b\u304b\u308f\u3089\u305a\u3001\u6f5c\u5728\u7684\u306a\u554f\u984c\u306f\u672a\u89e3\u6c7a\u306e\u307e\u307e\u3067\u3042\u308a\u3001\u305d\u306e\u4fe1\u983c\u6027\u3068\u5b9f\u7528\u6027\u306e\u7bc4\u56f2\u3092\u640d\u306a\u3046\u3002 \u63d0\u6848\u624b\u6cd5\u306f, LLM-as-a-Judge\u306b\u304a\u3051\u308b\u5404\u7a2e\u985e\u306e\u30d0\u30a4\u30a2\u30b9\u3092\u5b9a\u91cf\u5316\u3057, \u89e3\u6790\u3059\u308b\u81ea\u52d5\u30d0\u30a4\u30a2\u30b9\u5b9a\u91cf\u5316\u30d5\u30ec\u30fc\u30e0\u30ef\u30fc\u30af\u3067\u3042\u308b\u3002 \u5f53\u793e\u306e\u4f5c\u696d\u306f\u3001\u3053\u308c\u3089\u306e\u554f\u984c\u306b\u5bfe\u51e6\u3059\u308b\u30b9\u30c6\u30fc\u30af\u30db\u30eb\u30c0\u306e\u5fc5\u8981\u6027\u3092\u5f37\u8abf\u3057\u3001LLM-as-a-Judge\u30a2\u30d7\u30ea\u30b1\u30fc\u30b7\u30e7\u30f3\u3067\u6ce8\u610f\u3092\u559a\u8d77\u3057\u307e\u3059\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2410.02736v1\">\u8ad6\u6587<\/a>\u00a0\u00a0<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2410.02736v1\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>\u00a0 \u00a0(Thu, 03 Oct 2024 17:53:30 GMT)<\/li>\n\n\n\n<li>\u6700\u8fd1\u3088\u304f\u4f7f\u308f\u308c\u3066\u3044\u308bLLM as a Judge\u3067\u751f\u3058\u308b\u30d0\u30a4\u30a2\u30b9\u306e\u6574\u7406\u3068\u5b9a\u91cf\u5316\u306b\u5bfe\u3059\u308b\u63d0\u6848\u3002\u300cWhile Claude-3.5 generally shows the greatest resilience to biases, our findings reveal that even highly proficient models can struggle.\u300d\u3068\u3044\u3046\u7d50\u679c\u306f\u8208\u5473\u6df1\u3044\u3002\uff08GPT-4o\u306fClaude 3.5\u3088\u308a\u7d50\u679c\u304c\u60aa\u304b\u3063\u305f\uff09<\/li>\n\n\n\n<li>\u30ea\u30dd\u30b8\u30c8\u30ea\u306f<a href=\"https:\/\/llm-judge-bias.github.io\/\">Justice or Prejudice? Quantifying Biases in LLM-as-a-Judge (llm-judge-bias.github.io)<\/a><\/li>\n<\/ul>\n\n\n\n<p><\/p>\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":[224,647],"class_list":["post-5549","post","type-post","status-publish","format-standard","hentry","category-arxiv","tag-llm-as-a-judge","tag-647"],"_links":{"self":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/5549","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=5549"}],"version-history":[{"count":0,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/5549\/revisions"}],"wp:attachment":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=5549"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=5549"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=5549"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}