{"id":4352,"date":"2024-01-29T05:03:00","date_gmt":"2024-01-28T20:03:00","guid":{"rendered":"https:\/\/devneko.jp\/wordpress\/?p=4352"},"modified":"2024-01-29T05:03:00","modified_gmt":"2024-01-28T20:03:00","slug":"in-context-learning%e3%81%ae%e5%88%86%e6%9e%90%e3%81%a8%e3%82%b5%e3%83%bc%e3%83%99%e3%82%a4","status":"publish","type":"post","link":"https:\/\/devneko.jp\/wordpress\/?p=4352","title":{"rendered":"In context learning\u306e\u5206\u6790\u3068\u30b5\u30fc\u30d9\u30a4"},"content":{"rendered":"\n<p>\u666e\u901a\u306b\u7528\u3044\u3089\u308c\u3066\u3044\u308bIn context learning\u3060\u304c\u3001\u305d\u306e\u52d5\u304d\u306b\u95a2\u3059\u308b\u5206\u6790\u3068\u691c\u7d22\u4f75\u7528\u306b\u95a2\u3059\u308b\u30b5\u30fc\u30d9\u30a4\u304c\u51fa\u3066\u3044\u305f\u3002AI \u4e8b\u696d\u8005\u30ac\u30a4\u30c9\u30e9\u30a4\u30f3\u6848\uff08<a href=\"https:\/\/www8.cao.go.jp\/cstp\/ai\/ai_senryaku\/7kai\/13gaidorain.pdf\">13gaidorain.pdf (cao.go.jp)<\/a>\uff09\u3067\u30b3\u30f3\u30c6\u30ad\u30b9\u30c8\u5185\u5b66\u7fd2\u3068\u547c\u3070\u308c\u3066\u3044\u308b\u3082\u306e\u3067\u3001\u306a\u3093\u3067\u3053\u3093\u306a\u3053\u3068\u304c\u3067\u304d\u308b\u306e\u304b\u306e\u89e3\u6790\u306f\u9032\u3093\u3067\u304d\u3066\u3044\u308b\uff08<a href=\"https:\/\/devneko.jp\/wordpress\/?p=3396\">What and How does In-Context Learning Learn? Bayesian Model Averaging, Parameterization, and Generalization \u2013 arXiv\u6700\u65b0\u8ad6\u6587\u306e\u7d39\u4ecb (devneko.jp)<\/a>\u3001<a href=\"https:\/\/devneko.jp\/wordpress\/?p=4016\">When Do Prompting and Prefix-Tuning Work? A Theory of Capabilities and Limitations \u2013 arXiv\u6700\u65b0\u8ad6\u6587\u306e\u7d39\u4ecb (devneko.jp)<\/a>\uff09\u3082\u306e\u306e\u3002\u3084\u3063\u3071\u308a\u4e0d\u601d\u8b70\u3060\u306a\u30fc\u3068\u601d\u3046\u3002<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>In-Context Language Learning: Arhitectures and Algorithms\u00a0<\/strong>[73.9]<br>\u6211\u3005\u306f\u3001\u6587\u8108\u8a00\u8a9e\u5b66\u7fd2(ICLL)\u306b\u304a\u3044\u3066\u3001\u79c1\u305f\u3061\u304c\u7528\u8a9e\u3059\u308b\u65b0\u3057\u3044\u30e2\u30c7\u30eb\u554f\u984c\u7fa4(\u82f1\u8a9e\u7248)\u306e\u30ec\u30f3\u30ba\u3092\u901a\u3057\u3066ICL\u3092\u7814\u7a76\u3059\u308b\u3002 \u6211\u3005\u306f,\u901a\u5e38\u306eICLL\u30bf\u30b9\u30af\u306b\u304a\u3044\u3066,\u591a\u7a2e\u591a\u69d8\u306a\u30cb\u30e5\u30fc\u30e9\u30eb\u30b7\u30fc\u30b1\u30f3\u30b9\u30e2\u30c7\u30eb\u3092\u8a55\u4fa1\u3059\u308b\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2401.12973v1\">\u8ad6\u6587<\/a>\u00a0\u00a0<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2401.12973v1\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>\u00a0 \u00a0(Tue, 23 Jan 2024 18:59:21 GMT)<\/li>\n\n\n\n<li>\u300cthis paper provides evidence supporting the hypothesis that real language models can in-context learn using known learning algorithms.\u300d\u3068\u306e\u3053\u3068\u3002<\/li>\n\n\n\n<li>\u300cTransformers significantly outperform neural sequence models with recurrent or convolutional representations on ICLL tasks.\u300d\u3068\u3082\u66f8\u304b\u308c\u3066\u3044\u3066\uff08ICLL\uff1d\u00a0in-context language learning\u3001\u672a\u77e5\u306e\u5f62\u5f0f\u8a00\u8a9e\u306b\u5bfe\u3059\u308b\u63a8\u8ad6\u3067\u30aa\u30f3\u30b6\u30d5\u30e9\u30a4\u3067\u5b66\u7fd2\u3057\u306a\u3044\u3068\u5bfe\u5fdc\u3067\u304d\u306a\u3044\uff09Transformer\u306e\u5f8c\u7d99\u3092\u72d9\u3063\u305f\u30e2\u30c7\u30eb\u306fTransformer\u306b\u5339\u6575\u3067\u304d\u3066\u3044\u306a\u3044\u3002<\/li>\n<\/ul>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>In-context Learning with Retrieved Demonstrations for Language Models: A Survey\u00a0<\/strong>[22.4]<br>\u30a4\u30f3\u30b3\u30f3\u30c6\u30af\u30b9\u30c8\u5b66\u7fd2\u8005(ICL)\u306f\u5165\u529b\u30b3\u30f3\u30c6\u30ad\u30b9\u30c8\u3067\u306e\u30c7\u30e2\u3092\u5c11\u3057\u3060\u3051\u884c\u3046\u3060\u3051\u3067\u3001\u65b0\u3057\u3044\u30bf\u30b9\u30af\u306b\u9069\u5fdc\u3067\u304d\u308b\u3002 \u6700\u8fd1\u306e\u958b\u767a\u3067\u306f\u3001\u56fa\u5b9a\u3055\u308c\u305f\u4e00\u9023\u306e\u30c7\u30e2\u3092\u4f7f\u3046\u4ee3\u308f\u308a\u306b\u3001\u5404\u5165\u529b\u30af\u30a8\u30ea\u306b\u5408\u308f\u305b\u305f\u30c7\u30e2\u3092\u691c\u7d22\u3059\u308b\u3002 \u672c\u7a3f\u3067\u306f,\u691c\u7d22\u30e2\u30c7\u30eb,\u691c\u7d22\u8a13\u7df4\u624b\u9806,\u63a8\u8ad6\u30a2\u30eb\u30b4\u30ea\u30ba\u30e0\u306e\u7570\u306a\u308b\u8a2d\u8a08\u9078\u629e\u306b\u3064\u3044\u3066\u8ad6\u3058,\u6bd4\u8f03\u3059\u308b\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2401.11624v1\">\u8ad6\u6587<\/a>\u00a0\u00a0<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2401.11624v1\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>\u00a0 \u00a0(Sun, 21 Jan 2024 23:34:42 GMT)<\/li>\n\n\n\n<li>\u3053\u3061\u3089\u306f\u4e0e\u3048\u308b\u60c5\u5831\u3092\u5f97\u308b\u624b\u6cd5\u3092\u4e2d\u5fc3\u3068\u3057\u305f\u30b5\u30fc\u30d9\u30a4<\/li>\n\n\n\n<li>\u5b9f\u7528\u7684\u306b\u4f7f\u3046\u305f\u3081\u306b\u53c2\u8003\u306b\u306a\u308b\u60c5\u5831<\/li>\n<\/ul>\n","protected":false},"excerpt":{"rendered":"<p>\u666e\u901a\u306b\u7528\u3044\u3089\u308c\u3066\u3044\u308bIn context learning\u3060\u304c\u3001\u305d\u306e\u52d5\u304d\u306b\u95a2\u3059\u308b\u5206\u6790\u3068\u691c\u7d22\u4f75\u7528\u306b\u95a2\u3059\u308b\u30b5\u30fc\u30d9\u30a4\u304c\u51fa\u3066\u3044\u305f\u3002AI \u4e8b\u696d\u8005\u30ac\u30a4\u30c9\u30e9\u30a4\u30f3\u6848\uff0813gaidorain.pdf (cao.go.jp)\uff09\u3067\u30b3\u30f3\u30c6\u30ad\u30b9\u30c8 &hellip; <a href=\"https:\/\/devneko.jp\/wordpress\/?p=4352\" class=\"more-link\"><span class=\"screen-reader-text\">&#8220;In context learning\u306e\u5206\u6790\u3068\u30b5\u30fc\u30d9\u30a4&#8221; \u306e<\/span>\u7d9a\u304d\u3092\u8aad\u3080<\/a><\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[2],"tags":[200,387],"class_list":["post-4352","post","type-post","status-publish","format-standard","hentry","category-arxiv","tag-in-context-learning","tag-survey"],"_links":{"self":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/4352","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=4352"}],"version-history":[{"count":0,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/4352\/revisions"}],"wp:attachment":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=4352"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=4352"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=4352"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}