{"id":7380,"date":"2025-09-09T05:27:00","date_gmt":"2025-09-08T20:27:00","guid":{"rendered":"https:\/\/devneko.jp\/wordpress\/?p=7380"},"modified":"2025-09-06T20:30:13","modified_gmt":"2025-09-06T11:30:13","slug":"loong-synthesize-long-chain-of-thoughts-at-scale-through-verifiers","status":"publish","type":"post","link":"https:\/\/devneko.jp\/wordpress\/?p=7380","title":{"rendered":"Loong: Synthesize Long Chain-of-Thoughts at Scale through Verifiers"},"content":{"rendered":"\n<ul class=\"wp-block-list\">\n<li><strong>Loong: Synthesize Long Chain-of-Thoughts at Scale through Verifiers\u00a0<\/strong>[103.4]<br>\u30b9\u30b1\u30fc\u30e9\u30d6\u30eb\u306a\u5408\u6210\u30c7\u30fc\u30bf\u751f\u6210\u3068\u691c\u8a3c\u306e\u305f\u3081\u306e\u30aa\u30fc\u30d7\u30f3\u30bd\u30fc\u30b9\u306e\u30d5\u30ec\u30fc\u30e0\u30ef\u30fc\u30af\u3067\u3042\u308bLoong Project\u3092\u7d39\u4ecb\u3057\u307e\u3059\u3002 LoongBench\u306f\u300112\u306e\u30c9\u30e1\u30a4\u30f3\u306b\u307e\u305f\u304c\u308b8,729\u306e\u4eba\u70ba\u7684\u306a\u30b5\u30f3\u30d7\u30eb\u3092\u542b\u3080\u3001\u30ad\u30e5\u30ec\u30fc\u30c8\u3055\u308c\u305f\u30b7\u30fc\u30c9\u30c7\u30fc\u30bf\u30bb\u30c3\u30c8\u3067\u3042\u308b\u3002 LoongEnv\u306f\u30e2\u30b8\u30e5\u30e9\u30fc\u5408\u6210\u30c7\u30fc\u30bf\u751f\u6210\u74b0\u5883\u3067\u3042\u308a\u3001\u65b0\u3057\u3044\u8cea\u554f\u5fdc\u7b54\u30b3\u30fc\u30c9\u306e\u30c8\u30ea\u30d7\u30eb\u3092\u751f\u6210\u3059\u308b\u8907\u6570\u306e\u30d7\u30ed\u30f3\u30d7\u30c8\u6226\u7565\u3092\u30b5\u30dd\u30fc\u30c8\u3059\u308b\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2509.03059v1\">\u8ad6\u6587<\/a>\u00a0\u00a0<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2509.03059v1\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>\u00a0 \u00a0(Wed, 03 Sep 2025 06:42:40 GMT)<\/li>\n\n\n\n<li>\u300cOur contributions are fourfold: (1) LOONGBENCH, a seed dataset of 8,729 examples across 12 reasoning- intensive domains with executable code and verified answers; (2) LOONGENV, a flexible environment enabling diverse synthetic data generation strategies; (3) comprehensive benchmarking of open-source and proprietary models to assess domain generalization; and (4) in-depth analysis of generated data quality in terms of correctness, diversity, and complexity. Together, these components form a cohesive framework for studying alignment at scale.\u300d\u3068\u3001\u5408\u6210\u30c7\u30fc\u30bf\u306b\u95a2\u3059\u308b\u30d5\u30ec\u30fc\u30e0\u30ef\u30fc\u30af\u306e\u63d0\u6848\u3002\u5408\u6210\u30c7\u30fc\u30bf\u6d3b\u7528\u306f\u9ad8\u6027\u80fd\u30e2\u30c7\u30eb\u3092\u69cb\u7bc9\u3059\u308b\u306b\u3042\u305f\u308a\u57fa\u672c\u7684\u306a\u30a2\u30d7\u30ed\u30fc\u30c1\u3068\u306a\u3063\u3066\u304a\u308a\u3001\u3053\u306e\u624b\u306e\u30d5\u30ec\u30fc\u30e0\u30ef\u30fc\u30af\u306f\u3042\u308a\u304c\u305f\u3044\u3002<\/li>\n\n\n\n<li>\u30ea\u30dd\u30b8\u30c8\u30ea\u306f<a href=\"https:\/\/github.com\/camel-ai\/loong\">GitHub &#8211; camel-ai\/loong: \ud83d\udc09 Loong: Synthesize Long CoTs at Scale through Verifiers.<\/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":[390],"class_list":["post-7380","post","type-post","status-publish","format-standard","hentry","category-arxiv","tag-synthetic-data"],"_links":{"self":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/7380","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=7380"}],"version-history":[{"count":1,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/7380\/revisions"}],"predecessor-version":[{"id":7381,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/7380\/revisions\/7381"}],"wp:attachment":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=7380"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=7380"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=7380"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}