{"id":8031,"date":"2026-01-07T05:10:00","date_gmt":"2026-01-06T20:10:00","guid":{"rendered":"https:\/\/devneko.jp\/wordpress\/?p=8031"},"modified":"2026-01-04T11:16:52","modified_gmt":"2026-01-04T02:16:52","slug":"enhancing-llm-planning-capabilities-through-intrinsic-self-critique","status":"publish","type":"post","link":"https:\/\/devneko.jp\/wordpress\/?p=8031","title":{"rendered":"Enhancing LLM Planning Capabilities through Intrinsic Self-Critique"},"content":{"rendered":"\n<ul class=\"wp-block-list\">\n<li><strong>Enhancing LLM Planning Capabilities through Intrinsic Self-Critique\u00a0<\/strong>[34.8]<br>\u691c\u8a3c\u5668\u306a\u3069\u306e\u5916\u90e8\u30bd\u30fc\u30b9\u3092\u4f7f\u308f\u305a\u306b\u3001\u672c\u8cea\u7684\u306a\u81ea\u5df1\u6279\u5224\u3092\u901a\u3058\u3066\u30c7\u30fc\u30bf\u30bb\u30c3\u30c8\u3092\u8a08\u753b\u3059\u308b\u969b\u306e\u9855\u8457\u306a\u6027\u80fd\u5411\u4e0a\u3092\u793a\u3059\u3002 \u81ea\u5df1\u6279\u5224\u304c\u8a08\u753b\u306e\u30d1\u30d5\u30a9\u30fc\u30de\u30f3\u30b9\u3092\u5927\u5e45\u306b\u5411\u4e0a\u3055\u305b\u308b\u65b9\u6cd5\u306b\u3064\u3044\u3066\u8aac\u660e\u3059\u308b\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2512.24103v1\">\u8ad6\u6587<\/a>\u00a0\u00a0<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2512.24103v1\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>\u00a0 \u00a0(Tue, 30 Dec 2025 09:23:25 GMT)<\/li>\n\n\n\n<li>\u300cEach iteration of the self-improvement mechanism comprises two key steps: i) plan generation and ii) self-critiquing, aimed at iteratively refining LLM outputs. In step i), the LLM generates a plan (symbolized by a map) based on a prompt incorporating domain-specific knowledge and instructions (symbolized by the treasure chest). Step ii) involves a self-critique mechanism where the LLM evaluates its own performance, providing correctness assessments and justifications, again leveraging domain knowledge.\u300d\u3068\u81ea\u5df1\u6279\u5224\u306b\u3088\u308b\u6539\u5584\u624b\u6cd5\u306e\u63d0\u6848\u3002<\/li>\n\n\n\n<li>\u305d\u308c\u306a\u308a\u306b\u4f7f\u308f\u308c\u308b\u30c6\u30af\u30cb\u30c3\u30af\u3067\u3042\u308b\u3068\u306f\u601d\u3046\u306e\u3060\u304c\u3001\u30a4\u30c6\u30ec\u30fc\u30b7\u30e7\u30f3\u3092\u542b\u3081\u3057\u3063\u304b\u308a\u3068\u691c\u8a3c\u3055\u308c\u3066\u3044\u3066\u3068\u3066\u3082\u53c2\u8003\u306b\u306a\u308b\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":[356,636],"class_list":["post-8031","post","type-post","status-publish","format-standard","hentry","category-arxiv","tag-self-x","tag-636"],"_links":{"self":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/8031","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=8031"}],"version-history":[{"count":1,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/8031\/revisions"}],"predecessor-version":[{"id":8032,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/8031\/revisions\/8032"}],"wp:attachment":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=8031"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=8031"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=8031"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}