{"id":5511,"date":"2024-10-02T06:26:00","date_gmt":"2024-10-01T21:26:00","guid":{"rendered":"https:\/\/devneko.jp\/wordpress\/?p=5511"},"modified":"2024-10-02T06:26:00","modified_gmt":"2024-10-01T21:26:00","slug":"pretraining-data-detection-for-large-language-models-a-divergence-based-calibration-method","status":"publish","type":"post","link":"https:\/\/devneko.jp\/wordpress\/?p=5511","title":{"rendered":"Pretraining Data Detection for Large Language Models: A Divergence-based Calibration Method\u00a0"},"content":{"rendered":"\n<ul class=\"wp-block-list\">\n<li><strong>Pretraining Data Detection for Large Language Models: A Divergence-based Calibration Method&nbsp;<\/strong>[108.6]<br>\u672c\u7814\u7a76\u3067\u306f,\u4e71\u6570\u304b\u3089\u767a\u6563\u3059\u308b\u6982\u5ff5\u306b\u89e6\u767a\u3055\u308c\u305f\u504f\u5dee\u306b\u57fa\u3065\u304f\u30ad\u30e3\u30ea\u30d6\u30ec\u30fc\u30b7\u30e7\u30f3\u624b\u6cd5\u3092\u5c0e\u5165\u3057,\u30d7\u30ea\u30c8\u30ec\u30fc\u30cb\u30f3\u30b0\u30c7\u30fc\u30bf\u691c\u51fa\u306e\u305f\u3081\u306e\u30c8\u30fc\u30af\u30f3\u78ba\u7387\u306e\u30ad\u30e3\u30ea\u30d6\u30ec\u30fc\u30b7\u30e7\u30f3\u3092\u884c\u3046\u3002 \u6211\u3005\u306f,\u4e2d\u56fd\u8a9e\u30c6\u30ad\u30b9\u30c8\u4e0a\u3067\u306eLLM\u306e\u691c\u51fa\u624b\u6cd5\u306e\u6027\u80fd\u3092\u8a55\u4fa1\u3059\u308b\u305f\u3081\u306b,\u4e2d\u56fd\u8a9e\u306e\u30d9\u30f3\u30c1\u30de\u30fc\u30af\u3067\u3042\u308bPatentMIA\u3092\u958b\u767a\u3057\u305f\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2409.14781v1\">\u8ad6\u6587<\/a>&nbsp;&nbsp;<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2409.14781v1\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>&nbsp; &nbsp;(Mon, 23 Sep 2024 07:55:35 GMT)<\/li>\n\n\n\n<li>\u4e8b\u524d\u5b66\u7fd2\u306b\u4f55\u304c\u4f7f\u308f\u308c\u305f\u304b\u3092\u691c\u77e5\u3059\u308b\u30bf\u30b9\u30afpretraining data detection\u306b\u95a2\u3059\u308b\u624b\u6cd5DC-PDD \u304a\u3088\u3073\u30d9\u30f3\u30c1\u30de\u30fc\u30af\u306e\u63d0\u6848\u3002\u300cThe pretraining data detection problem can be viewed as an instance of the membership inference attack (MIA) task (Shokri et al , 2017), where the primary objective is to determine if a particular text was part of a target LLM\u2019s training corpus.\u300d<\/li>\n\n\n\n<li>DC-PDD computes the divergence between the token probability distribution and the token frequency distribution for detection.\u3068\u306e\u3053\u3068\u3002<\/li>\n\n\n\n<li>\u30ea\u30dd\u30b8\u30c8\u30ea\u306f<a href=\"https:\/\/github.com\/zhang-wei-chao\/DC-PDD\">GitHub &#8211; zhang-wei-chao\/DC-PDD<\/a><\/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":[223,242],"class_list":["post-5511","post","type-post","status-publish","format-standard","hentry","category-arxiv","tag-llm","tag-membership"],"_links":{"self":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/5511","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=5511"}],"version-history":[{"count":0,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/5511\/revisions"}],"wp:attachment":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=5511"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=5511"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=5511"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}