{"id":7866,"date":"2025-12-09T05:27:00","date_gmt":"2025-12-08T20:27:00","guid":{"rendered":"https:\/\/devneko.jp\/wordpress\/?p=7866"},"modified":"2025-12-06T16:56:38","modified_gmt":"2025-12-06T07:56:38","slug":"thetaevolve-test-time-learning-on-open-problems","status":"publish","type":"post","link":"https:\/\/devneko.jp\/wordpress\/?p=7866","title":{"rendered":"ThetaEvolve: Test-time Learning on Open Problems"},"content":{"rendered":"\n<ul class=\"wp-block-list\">\n<li><strong>ThetaEvolve: Test-time Learning on Open Problems\u00a0<\/strong>[110.6]<br>In-context Learning\u3068Reinforcement Learning(RL)\u306e\u4e21\u65b9\u3092\u30c6\u30b9\u30c8\u6642\u306b\u52b9\u7387\u7684\u306b\u30b9\u30b1\u30fc\u30eb\u3059\u308b\u305f\u3081\u306b\u3001AlphaEvolve\u3092\u5358\u7d14\u5316\u3057\u62e1\u5f35\u3059\u308b\u30aa\u30fc\u30d7\u30f3\u30bd\u30fc\u30b9\u306e\u30d5\u30ec\u30fc\u30e0\u30ef\u30fc\u30af\u3067\u3042\u308bThetaEvolve\u3092\u7d39\u4ecb\u3057\u307e\u3059\u3002 \u30c6\u30b9\u30c8\u6642\u306bRL\u3092\u4f7f\u7528\u3059\u308bThetaEvolve\u306f\u3001\u63a8\u8ad6\u306e\u307f\u306e\u30d9\u30fc\u30b9\u30e9\u30a4\u30f3\u3088\u308a\u3082\u4e00\u8cab\u3057\u3066\u512a\u308c\u3066\u3044\u307e\u3059\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2511.23473v1\">\u8ad6\u6587<\/a>\u00a0\u00a0<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2511.23473v1\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>\u00a0 \u00a0(Fri, 28 Nov 2025 18:58:14 GMT)<\/li>\n\n\n\n<li>\u300cWe introduce ThetaEvolve, an open-source framework that simplifies and extends AlphaEvolve to efficiently scale both in-context learning and Reinforcement Learning (RL) at test time, allowing models to continually learn from their experiences in improving open optimization problems. ThetaEvolve features a single LLM, a large pro- gram database for enhanced exploration, batch sampling for higher throughput, lazy penalties to discourage stagnant outputs, and optional reward shaping for stable training signals, etc.\u300d\u3068OSS\u7248AlphaEvolve\u7684\u306a\u7814\u7a76\u3002\u300c(2) Surprisingly, we show that when scaling test-time compute with ThetaEvolve, a single open-source 8B model, DeepSeek-R1-0528-Qwen3-8B (DeepSeek-AI, 2025), can improve the best-known bounds of two open problems considered in AlphaEvolve\u300d\u3068\u52b9\u679c\u3092\u78ba\u8a8d\u3057\u3066\u3044\u308b\u3002<\/li>\n\n\n\n<li>\u30ea\u30dd\u30b8\u30c8\u30ea\u306f<a href=\"https:\/\/github.com\/ypwang61\/ThetaEvolve\">GitHub &#8211; ypwang61\/ThetaEvolve: ThetaEvolve: Test-time Learning on Open Problems<\/a><\/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":[200,576],"class_list":["post-7866","post","type-post","status-publish","format-standard","hentry","category-arxiv","tag-in-context-learning","tag-576"],"_links":{"self":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/7866","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=7866"}],"version-history":[{"count":1,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/7866\/revisions"}],"predecessor-version":[{"id":7867,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/7866\/revisions\/7867"}],"wp:attachment":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=7866"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=7866"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=7866"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}