{"id":6367,"date":"2025-03-04T03:43:00","date_gmt":"2025-03-03T18:43:00","guid":{"rendered":"https:\/\/devneko.jp\/wordpress\/?p=6367"},"modified":"2025-03-04T03:43:00","modified_gmt":"2025-03-03T18:43:00","slug":"continuous-diffusion-model-for-language-modeling-energy-based-diffusion-language-models-for-text-generation","status":"publish","type":"post","link":"https:\/\/devneko.jp\/wordpress\/?p=6367","title":{"rendered":"Continuous Diffusion Model for Language Modeling, Energy-Based Diffusion Language Models for Text Generation\u00a0"},"content":{"rendered":"\n<ul class=\"wp-block-list\">\n<li><strong>Continuous Diffusion Model for Language Modeling\u00a0<\/strong>[57.4]<br>\u96e2\u6563\u30c7\u30fc\u30bf\u306b\u5bfe\u3059\u308b\u65e2\u5b58\u306e\u9023\u7d9a\u62e1\u6563\u30e2\u30c7\u30eb\u306f\u3001\u96e2\u6563\u7684\u30a2\u30d7\u30ed\u30fc\u30c1\u3068\u6bd4\u8f03\u3057\u3066\u6027\u80fd\u304c\u9650\u3089\u308c\u3066\u3044\u308b\u3002 \u672c\u7a3f\u3067\u306f,\u4e0b\u5c64\u306e\u5206\u985e\u5206\u5e03\u306e\u5e7e\u4f55\u5b66\u3092\u7d44\u307f\u8fbc\u3093\u3060\u8a00\u8a9e\u30e2\u30c7\u30ea\u30f3\u30b0\u306e\u305f\u3081\u306e\u9023\u7d9a\u62e1\u6563\u30e2\u30c7\u30eb\u3092\u63d0\u6848\u3059\u308b\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2502.11564v1\">\u8ad6\u6587<\/a>\u00a0\u00a0<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2502.11564v1\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>\u00a0 \u00a0(Mon, 17 Feb 2025 08:54:29 GMT)<\/li>\n\n\n\n<li>AR\u30e2\u30c7\u30eb\u306b\u5339\u6575\u3059\u308bRiemannian Diffusion Language Model (RDLM),\u306e\u63d0\u6848\u3002<\/li>\n\n\n\n<li>\u30ea\u30dd\u30b8\u30c8\u30ea\u306f<a href=\"https:\/\/github.com\/harryjo97\/RDLM\">https:\/\/github.com\/harryjo97\/RDLM<\/a><\/li>\n\n\n\n<li>\u753b\u50cf\u3067\u306fDiffusion Model \u2192 Autoregressive model\u3068\u3044\u3046\u6d41\u308c\u3082\u3042\u308a\u3064\u3064\u3001\u8a00\u8a9e\u3067\u306fDiffusion Model\u3092\u4f7f\u3046<a href=\"https:\/\/www.inceptionlabs.ai\/news\">Inception Labs<\/a>, <a href=\"https:\/\/chat.inceptionlabs.ai\/\">Mercury Coder<\/a>\u304c\u8a71\u984c\u306b\u306a\u3063\u3066\u3044\u308b\u306e\u304c\u9762\u767d\u3044\u3002<\/li>\n<\/ul>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>Energy-Based Diffusion Language Models for Text Generation\u00a0<\/strong>[126.2]<br>\u30a8\u30cd\u30eb\u30ae\u30fc\u30d9\u30fc\u30b9\u62e1\u6563\u8a00\u8a9e\u30e2\u30c7\u30eb(Energy-based Diffusion Language Model, EDLM)\u306f\u3001\u62e1\u6563\u30b9\u30c6\u30c3\u30d7\u3054\u3068\u306b\u5168\u30b7\u30fc\u30b1\u30f3\u30b9\u30ec\u30d9\u30eb\u3067\u52d5\u4f5c\u3059\u308b\u30a8\u30cd\u30eb\u30ae\u30fc\u30d9\u30fc\u30b9\u30e2\u30c7\u30eb\u3067\u3042\u308b\u3002 \u6211\u3005\u306e\u30d5\u30ec\u30fc\u30e0\u30ef\u30fc\u30af\u306f\u3001\u65e2\u5b58\u306e\u62e1\u6563\u30e2\u30c7\u30eb\u3088\u308a\u30821.3$times$\u306e\u30b5\u30f3\u30d7\u30ea\u30f3\u30b0\u30b9\u30d4\u30fc\u30c9\u30a2\u30c3\u30d7\u3092\u63d0\u4f9b\u3059\u308b\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2410.21357v3\">\u8ad6\u6587<\/a>\u00a0\u00a0<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2410.21357v3\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>\u00a0 \u00a0(Fri, 28 Feb 2025 08:41:03 GMT)<\/li>\n\n\n\n<li>\u3053\u3061\u3089\u3082\u300cThrough experiments on both small and large language modeling benchmarks, EDLM demonstrates state-of-the-art performance among diffusion models and approaches the quality of autoregressive models, while offering significant sampling speedup.\u300d\u3092\u4e3b\u5f35\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":[114,223],"class_list":["post-6367","post","type-post","status-publish","format-standard","hentry","category-arxiv","tag-diffusion-model","tag-llm"],"_links":{"self":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/6367","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=6367"}],"version-history":[{"count":0,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/6367\/revisions"}],"wp:attachment":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=6367"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=6367"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=6367"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}