{"id":7905,"date":"2025-12-19T05:10:00","date_gmt":"2025-12-18T20:10:00","guid":{"rendered":"https:\/\/devneko.jp\/wordpress\/?p=7905"},"modified":"2025-12-13T07:14:32","modified_gmt":"2025-12-12T22:14:32","slug":"opv-outcome-based-process-verifier-for-efficient-long-chain-of-thought-verification","status":"publish","type":"post","link":"https:\/\/devneko.jp\/wordpress\/?p=7905","title":{"rendered":"OPV: Outcome-based Process Verifier for Efficient Long Chain-of-Thought Verification\u00a0"},"content":{"rendered":"\n<ul class=\"wp-block-list\">\n<li><strong>OPV: Outcome-based Process Verifier for Efficient Long Chain-of-Thought Verification\u00a0<\/strong>[91.2]<br>\u672c\u7a3f\u3067\u306f\u3001\u9577\u3044\u601d\u8003\u306e\u9023\u9396\u304b\u3089\u8981\u7d04\u3055\u308c\u305f\u7d50\u679c\u306e\u5408\u7406\u5316\u904e\u7a0b\u3092\u691c\u8a3c\u3059\u308b\u3001\u30a2\u30a6\u30c8\u30ab\u30e0\u30d9\u30fc\u30b9\u30d7\u30ed\u30bb\u30b9\u691c\u8a3c(OPV)\u3092\u63d0\u6848\u3059\u308b\u3002 OPV \u306f 76.3 \u3068\u6bd4\u8f03\u3057\u3066 F1 \u30b9\u30b3\u30a2\u304c 83.1 \u306e Qwen3-Max-Preview \u306a\u3069,\u306f\u308b\u304b\u306b\u5927\u304d\u306a\u30aa\u30fc\u30d7\u30f3\u30bd\u30fc\u30b9\u30e2\u30c7\u30eb\u3088\u308a\u3082\u512a\u308c\u3066\u3044\u307e\u3059\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2512.10756v1\">\u8ad6\u6587<\/a>\u00a0\u00a0<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2512.10756v1\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>\u00a0 \u00a0(Thu, 11 Dec 2025 15:47:38 GMT)<\/li>\n\n\n\n<li>\u300cWe introduced the Outcome-based Process Verifier (OPV), which bridges outcome and process verification by operating on summarized solutions from long CoTs. Through an iterative active learning framework with expert annotations, OPV progressively improves its verification capabilities while minimizing annotation costs.\u300d\u3068CoT\u7684\u306a\u63a8\u8ad6\u904e\u7a0b\u3092\u691c\u8a3c\u3059\u308b\u30a2\u30d7\u30ed\u30fc\u30c1\u306e\u63d0\u6848\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":[59,84],"class_list":["post-7905","post","type-post","status-publish","format-standard","hentry","category-arxiv","tag-chain-of-thought","tag-critic"],"_links":{"self":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/7905","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=7905"}],"version-history":[{"count":1,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/7905\/revisions"}],"predecessor-version":[{"id":7906,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/7905\/revisions\/7906"}],"wp:attachment":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=7905"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=7905"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=7905"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}