{"id":6480,"date":"2025-03-24T04:55:00","date_gmt":"2025-03-23T19:55:00","guid":{"rendered":"https:\/\/devneko.jp\/wordpress\/?p=6480"},"modified":"2025-03-24T04:55:00","modified_gmt":"2025-03-23T19:55:00","slug":"empowering-llms-in-decision-games-through-algorithmic-data-synthesis","status":"publish","type":"post","link":"https:\/\/devneko.jp\/wordpress\/?p=6480","title":{"rendered":"Empowering LLMs in Decision Games through Algorithmic Data Synthesis"},"content":{"rendered":"\n<ul class=\"wp-block-list\">\n<li><strong>Empowering LLMs in Decision Games through Algorithmic Data Synthesis\u00a0<\/strong>[29.1]<br>\u610f\u601d\u6c7a\u5b9a\u30b2\u30fc\u30e0\u306f\u3001\u5927\u898f\u6a21\u8a00\u8a9e\u30e2\u30c7\u30eb\u306e\u63a8\u8ad6\u80fd\u529b\u3092\u8a55\u4fa1\u3057\u3001\u5f37\u5316\u3059\u308b\u305f\u3081\u306e\u7406\u60f3\u7684\u306a\u30b5\u30f3\u30c9\u30dc\u30c3\u30af\u30b9\u3068\u3057\u3066\u6a5f\u80fd\u3059\u308b\u3002 \u30c7\u30fc\u30bf\u5408\u6210\u6226\u7565\u3092\u8a2d\u8a08\u3057\u30012\u3064\u306e\u53e4\u5178\u30b2\u30fc\u30e0\u3001Doudizhu\u3068Go\u304b\u3089\u5e83\u7bc4\u56f2\u306e\u30aa\u30d5\u30e9\u30a4\u30f3\u30c7\u30fc\u30bf\u30bb\u30c3\u30c8\u3092\u30ad\u30e5\u30ec\u30fc\u30c8\u3059\u308b\u3002 \u6211\u3005\u306f\u3001\u3053\u306e\u30c7\u30fc\u30bf\u3092LLM\u30c8\u30ec\u30fc\u30cb\u30f3\u30b0\u306b\u52b9\u679c\u7684\u306b\u7d44\u307f\u8fbc\u3080\u305f\u3081\u306e\u4e00\u9023\u306e\u6280\u8853\u3092\u958b\u767a\u3057\u3001\u305d\u306e\u7d50\u679c\u3001Mastermind-Dou \u3068 Mastermind-Go \u3068\u3044\u30462\u3064\u306e\u65b0\u3057\u3044\u30a8\u30fc\u30b8\u30a7\u30f3\u30c8\u3092\u751f\u307f\u51fa\u3057\u305f\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2503.13980v1\">\u8ad6\u6587<\/a>\u00a0\u00a0<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2503.13980v1\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>\u00a0 \u00a0(Tue, 18 Mar 2025 07:30:29 GMT)<\/li>\n\n\n\n<li>\u4e00\u822c\u7684\u306b\u6570\u5b66\u3084\u30b3\u30fc\u30c9\u751f\u6210\u3092\u5bfe\u8c61\u306bLRM\u5316\u304c\u884c\u308f\u308c\u3066\u3044\u308b\u304c\u3053\u306e\u8ad6\u6587\u3067\u306f\u300cThrough a suite of our designed techniques in data collection and training, we have developed MasterMind agents, demonstrating commendable performance in both Doudizhu and Go.\u300d\u3068\u30b2\u30fc\u30e0\u304c\u5bfe\u8c61\u3002\u300cEmpirical experiments also serve to substantiate the potential of this approach in improving general reasoning capabilities of LLMs.\u300d\u3068\u3044\u3046\u306e\u304c\u3068\u3066\u3082\u8208\u5473\u6df1\u3044\u3002\u4eba\u9593\u3067\u3044\u3046\u3068\u3053\u308d\u306e\u300c\u8133\u306b\u3088\u3044\u3007\u3007\u300d\u7684\u306a\u30bf\u30b9\u30af\u304c\u3042\u308b\u306e\u3060\u308d\u3046\u304b\u3002\uff08\u3082\u3063\u3068\u3082\u6027\u80fd\u304c\u843d\u3061\u308b\u30bf\u30b9\u30af\u304c\u3042\u308b\u3053\u3068\u3082\u6307\u6458\u3055\u308c\u3066\u3044\u308b\u304c\u30fb\u30fb\u30fb\uff09<\/li>\n\n\n\n<li>\u30c7\u30fc\u30bf\u30bb\u30c3\u30c8\u304c\u516c\u958b\u3055\u308c\u3066\u3044\u308b\u3002<a href=\"https:\/\/huggingface.co\/datasets\/OpenDILabCommunity\/MasterMind\">OpenDILabCommunity\/MasterMind \u00b7 Datasets at Hugging Face<\/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":[121,223],"class_list":["post-6480","post","type-post","status-publish","format-standard","hentry","category-arxiv","tag-easy2hard","tag-llm"],"_links":{"self":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/6480","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=6480"}],"version-history":[{"count":0,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/6480\/revisions"}],"wp:attachment":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=6480"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=6480"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=6480"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}