{"id":5943,"date":"2024-12-24T05:17:00","date_gmt":"2024-12-23T20:17:00","guid":{"rendered":"https:\/\/devneko.jp\/wordpress\/?p=5943"},"modified":"2024-12-24T05:17:00","modified_gmt":"2024-12-23T20:17:00","slug":"thinkcite-improving-attributed-text-generation-with-self-guided-tree-search-and-progress-reward-modeling","status":"publish","type":"post","link":"https:\/\/devneko.jp\/wordpress\/?p=5943","title":{"rendered":"Think&amp;Cite\u3001RAG-Star"},"content":{"rendered":"\n<ul class=\"wp-block-list\">\n<li><strong>Think&amp;Cite: Improving Attributed Text Generation with Self-Guided Tree Search and Progress Reward Modeling&nbsp;<\/strong>[64.0]<br>\u5927\u578b\u8a00\u8a9e\u30e2\u30c7\u30eb(LLM)\u306f\u5e7b\u899a\u3092\u8d77\u3053\u3057\u3001\u4e8b\u5b9f\u7684\u306b\u8aa4\u3063\u305f\u60c5\u5831\u3092\u751f\u307f\u51fa\u3059\u50be\u5411\u306b\u3042\u308b\u3002 \u6211\u3005\u306fThink&amp;Cite\u3068\u547c\u3070\u308c\u308b\u65b0\u3057\u3044\u30d5\u30ec\u30fc\u30e0\u30ef\u30fc\u30af\u3092\u63d0\u6848\u3057\u3001\u691c\u7d22\u3068\u7d71\u5408\u3055\u308c\u305f\u591a\u6bb5\u968e\u63a8\u8ad6\u554f\u984c\u3068\u3057\u3066\u5c5e\u6027\u4ed8\u304d\u30c6\u30ad\u30b9\u30c8\u751f\u6210\u3092\u5b9a\u5f0f\u5316\u3059\u308b\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2412.14860v1\">\u8ad6\u6587<\/a>&nbsp;&nbsp;<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2412.14860v1\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>&nbsp; &nbsp;(Thu, 19 Dec 2024 13:55:48 GMT)<\/li>\n\n\n\n<li>\u30a8\u30d3\u30c7\u30f3\u30b9\u4ed8\u304d\u306e\u30c6\u30ad\u30b9\u30c8\u751f\u6210\u306e\u305f\u3081Self-Guided Monte Carlo Tree Search (SG-MCTS)\u3092\u63d0\u6848\u3002\u30e2\u30f3\u30c6\u30ab\u30eb\u30ed\u30c4\u30ea\u30fc\u3092\u4f7f\u3063\u3066\u6027\u80fd\u3092\u4e0a\u3052\u3088\u3046\u3068\u3044\u3046\u53d6\u308a\u7d44\u307f\u306f\u591a\u3044\u304c\u300cTo the best of our knowledge, we are the first to apply tree search algorithms to the task of attributed text generation.\u300d\u306f\u305d\u3046\u304b\u3082\u3057\u308c\u306a\u3044\u3002<\/li>\n\n\n\n<li>RAG\u306a\u3069\u3092\u4e0a\u56de\u308b\u6027\u80fd\u3092\u9054\u6210\u3068\u306e\u3053\u3068\u3002\u6709\u52b9\u306a\u624b\u6cd5\u306b\u601d\u3048\u308b\u3002<\/li>\n<\/ul>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>RAG-Star: Enhancing Deliberative Reasoning with Retrieval Augmented Verification and Refinement\u00a0<\/strong>[85.1]<br>\u65e2\u5b58\u306e\u5927\u898f\u6a21\u8a00\u8a9e\u30e2\u30c7\u30eb(LLM)\u306f\u3001\u4f8b\u5916\u7684\u306a\u554f\u984c\u89e3\u6c7a\u80fd\u529b\u3092\u793a\u3059\u304c\u3001\u8907\u96d1\u306a\u63a8\u8ad6\u30bf\u30b9\u30af\u306b\u82e6\u52b4\u3059\u308b\u53ef\u80fd\u6027\u304c\u3042\u308b\u3002 \u691c\u7d22\u60c5\u5831\u3092\u7d71\u5408\u3057\u305f\u65b0\u3057\u3044RAG\u624b\u6cd5\u3067\u3042\u308b RAG-Star \u3092\u63d0\u6848\u3059\u308b\u3002 Llama-3.1-8B-Instruct \u3068 GPT-4o \u3092\u4f75\u7528\u3057\u305f\u5b9f\u9a13\u306b\u3088\u308a,RAG-Star \u306f\u5f93\u6765\u306eRAG \u3068\u63a8\u7406\u6cd5\u3092\u8457\u3057\u304f\u4e0a\u56de\u3063\u3066\u3044\u308b\u3053\u3068\u304c\u793a\u3055\u308c\u305f\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2412.12881v1\">\u8ad6\u6587<\/a>\u00a0\u00a0<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2412.12881v1\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>\u00a0 \u00a0(Tue, 17 Dec 2024 13:05:36 GMT)<\/li>\n\n\n\n<li>\u300cRAG-Star employed Monte Carlo Tree Search to search intermediate sub-queries and corresponding answers. Moreover, RAG-Star introduced retrieval-augmented verification to evaluate the plausibility and consistency of the planned subqueries and answers based on a query-aware and an answer-aware reward.\u300d\u3068\u3053\u3061\u3089\u306fRAG\u306bMonte Carlo Tree Search\u3092\u7d44\u307f\u5408\u308f\u305b\u308b\u30bf\u30a4\u30d7\u306e\u5831\u544a<\/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":[42,324],"class_list":["post-5943","post","type-post","status-publish","format-standard","hentry","category-arxiv","tag-autonomous-agent","tag-rag"],"_links":{"self":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/5943","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=5943"}],"version-history":[{"count":0,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/5943\/revisions"}],"wp:attachment":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=5943"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=5943"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=5943"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}