{"id":6156,"date":"2025-02-04T05:51:00","date_gmt":"2025-02-03T20:51:00","guid":{"rendered":"https:\/\/devneko.jp\/wordpress\/?p=6156"},"modified":"2025-02-04T05:51:00","modified_gmt":"2025-02-03T20:51:00","slug":"think-smarter-not-harder-adaptive-reasoning-with-inference-aware-optimization","status":"publish","type":"post","link":"https:\/\/devneko.jp\/wordpress\/?p=6156","title":{"rendered":"Think Smarter not Harder: Adaptive Reasoning with Inference Aware Optimization"},"content":{"rendered":"\n<ul class=\"wp-block-list\">\n<li><strong>Think Smarter not Harder: Adaptive Reasoning with Inference Aware Optimization\u00a0<\/strong>[48.6]<br>\u672c\u7814\u7a76\u3067\u306f,\u30e2\u30c7\u30eb\u304c\u63a8\u8ad6\u4e88\u7b97\u306e\u5236\u7d04\u306b\u5bfe\u3057\u3066\u5b9f\u7528\u6027\u3068\u3057\u3066\u5b9a\u5f0f\u5316\u3059\u308b\u3053\u3068\u3067,\u63a8\u8ad6\u4e88\u7b97\u3092\u8a8d\u8b58\u3067\u304d\u308b\u3088\u3046\u306b\u3059\u308b\u624b\u6cd5\u3092\u63d0\u6848\u3059\u308b\u3002 \u7c21\u5358\u306b\u8a00\u3048\u3070\u3001IBPO\u3092\u901a\u3058\u3066\u5fae\u8abf\u6574\u3055\u308c\u305f\u30e2\u30c7\u30eb\u306f\u3001\u30af\u30a8\u30ea\u306e\u96e3\u3057\u3055\u3092\u7406\u89e3\u3057\u3001\u3088\u308a\u96e3\u3057\u3044\u3082\u306e\u306b\u63a8\u8ad6\u4e88\u7b97\u3092\u5272\u308a\u5f53\u3066\u308b\u3002 \u3053\u308c\u3089\u306e\u6539\u5584\u306f\u3001\u540c\u3058\u4e88\u7b97\u306e\u4e0b\u3067\u306e\u81ea\u5df1\u6574\u5408\u6027(self-consistency)\u306e\u7d042\u30c9\u30eb\u3067\u3042\u308b\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2501.17974v1\">\u8ad6\u6587<\/a>\u00a0\u00a0<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2501.17974v1\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>\u00a0 \u00a0(Wed, 29 Jan 2025 20:20:48 GMT)<\/li>\n\n\n\n<li><a href=\"https:\/\/devneko.jp\/wordpress\/?p=6110\">O1-Pruner: Length-Harmonizing Fine-Tuning for O1-Like Reasoning Pruning\u00a0 \u2013 arXiv\u6700\u65b0\u8ad6\u6587\u306e\u7d39\u4ecb<\/a>\u306b\u8fd1\u3044\u30e2\u30c1\u30d9\u30fc\u30b7\u30e7\u30f3\u3068\u601d\u308f\u308c\u308b\u63a8\u8ad6\u4e88\u7b97\u3092\u6c17\u306b\u3059\u308b\u30d5\u30ec\u30fc\u30e0\u30ef\u30fc\u30af\u306e\u63d0\u6848\u3002\u300cIn this work, we propose a way to allow models to be aware of inference budgets by formulating it as utility maximization with respect to an inference budget constraint, hence naming our algorithm Inference Budget-Constrained Policy Optimization (IBPO).\u300d<\/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":[298],"class_list":["post-6156","post","type-post","status-publish","format-standard","hentry","category-arxiv","tag-overthinking"],"_links":{"self":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/6156","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=6156"}],"version-history":[{"count":0,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/6156\/revisions"}],"wp:attachment":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=6156"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=6156"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=6156"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}