{"id":4919,"date":"2024-05-28T05:47:00","date_gmt":"2024-05-27T20:47:00","guid":{"rendered":"https:\/\/devneko.jp\/wordpress\/?p=4919"},"modified":"2024-05-28T05:47:00","modified_gmt":"2024-05-27T20:47:00","slug":"agent-planning-with-world-knowledge-model","status":"publish","type":"post","link":"https:\/\/devneko.jp\/wordpress\/?p=4919","title":{"rendered":"Agent Planning with World Knowledge Model"},"content":{"rendered":"\n<ul class=\"wp-block-list\">\n<li><strong>Agent Planning with World Knowledge Model\u00a0<\/strong>[88.5]<br>\u30a8\u30fc\u30b8\u30a7\u30f3\u30c8\u8a08\u753b\u3092\u5bb9\u6613\u306b\u3059\u308b\u305f\u3081\u306b\u30d1\u30e9\u30e1\u30c8\u30ea\u30c3\u30af\u4e16\u754c\u77e5\u8b58\u30e2\u30c7\u30eb(WKM)\u3092\u5c0e\u5165\u3059\u308b\u3002 \u6211\u3005\u306fWKM\u3092\u958b\u767a\u3057\u3001\u30b0\u30ed\u30fc\u30d0\u30eb\u306a\u8a08\u753b\u3068\u52d5\u7684\u72b6\u614b\u306e\u77e5\u8b58\u3092\u5c0e\u304f\u305f\u3081\u306b\u3001\u4e8b\u524d\u306e\u30bf\u30b9\u30af\u77e5\u8b58\u3092\u63d0\u4f9b\u3059\u308b\u3002 \u6211\u3005\u306f\u3001\u6211\u3005\u306eWKM\u304c\u8996\u899a\u969c\u5bb3\u8005\u306e\u8a66\u884c\u932f\u8aa4\u3068\u5e7b\u899a\u7684\u884c\u52d5\u306e\u554f\u984c\u3092\u52b9\u679c\u7684\u306b\u7de9\u548c\u3067\u304d\u308b\u3053\u3068\u3092\u793a\u3059\u305f\u3081\u306b\u5206\u6790\u3092\u884c\u3063\u305f\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2405.14205v1\">\u8ad6\u6587<\/a>\u00a0\u00a0<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2405.14205v1\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>\u00a0 \u00a0(Thu, 23 May 2024 06:03:19 GMT)<\/li>\n\n\n\n<li>World Knowledge Model\u304c\u8a08\u753b\u306b\u6709\u52b9\u3068\u306e\u3053\u3068\u3002\u305d\u308c\u81ea\u4f53\u306f\u7d0d\u5f97\u7684\u3067WKM\u3092\u5f97\u308b\u305f\u3081\u306b\u300cSpecifically, we first steer the agent model to synthesize task knowledge from the comparison between expert and sampled trajectories. Then we prompt it to summarize state knowledge for each planning step from expert trajectories and combine the previous and next actions to build a state knowledge base. Lastly, we integrate the generated knowledge into expert trajectories and train a WKM.\u300d\u3068\u3044\u3046\u624b\u9806\u3092\u3068\u308b\u3002\u3053\u306e\u624b\u306e\u8a2d\u8a08\u304c\u91cd\u8981\u306b\u306a\u3063\u3066\u3044\u308b\u3002<\/li>\n\n\n\n<li>\u30ea\u30dd\u30b8\u30c8\u30ea\u306f<a href=\"https:\/\/github.com\/zjunlp\/WKM\">https:\/\/github.com\/zjunlp\/WKM<\/a>\u3068\u306e\u3053\u3068\u3060\u304c\u3001\u73fe\u6642\u70b9\u3067\u306f\uff14\uff10\uff14<\/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":[42,636],"class_list":["post-4919","post","type-post","status-publish","format-standard","hentry","category-arxiv","tag-autonomous-agent","tag-636"],"_links":{"self":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/4919","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=4919"}],"version-history":[{"count":0,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/4919\/revisions"}],"wp:attachment":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=4919"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=4919"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=4919"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}