{"id":7845,"date":"2025-12-01T04:44:00","date_gmt":"2025-11-30T19:44:00","guid":{"rendered":"https:\/\/devneko.jp\/wordpress\/?p=7845"},"modified":"2025-11-29T16:50:34","modified_gmt":"2025-11-29T07:50:34","slug":"evo-memory-benchmarking-llm-agent-test-time-learning-with-self-evolving-memory","status":"publish","type":"post","link":"https:\/\/devneko.jp\/wordpress\/?p=7845","title":{"rendered":"Evo-Memory: Benchmarking LLM Agent Test-time Learning with Self-Evolving Memory\u00a0"},"content":{"rendered":"\n<ul class=\"wp-block-list\">\n<li><strong>Evo-Memory: Benchmarking LLM Agent Test-time Learning with Self-Evolving Memory\u00a0<\/strong>[89.7]<br>Evo-Memory\u306f\u3001\u5927\u898f\u6a21\u8a00\u8a9e\u30e2\u30c7\u30eb(LLM)\u30a8\u30fc\u30b8\u30a7\u30f3\u30c8\u3067\u81ea\u5df1\u9032\u5316\u578b\u30e1\u30e2\u30ea\u3092\u8a55\u4fa1\u3059\u308b\u305f\u3081\u306e\u3001\u30b9\u30c8\u30ea\u30fc\u30df\u30f3\u30b0\u30d9\u30f3\u30c1\u30de\u30fc\u30af\u3068\u30d5\u30ec\u30fc\u30e0\u30ef\u30fc\u30af\u3067\u3042\u308b\u3002 10\u4ee5\u4e0a\u306e\u4ee3\u8868\u7684\u306a\u30e1\u30e2\u30ea\u30e2\u30b8\u30e5\u30fc\u30eb\u3092\u8a55\u4fa1\u3057\u300110\u7a2e\u985e\u306e\u591a\u30bf\u30fc\u30f3\u30b4\u30fc\u30eb\u6307\u5411\u304a\u3088\u3073\u30b7\u30f3\u30b0\u30eb\u30bf\u30fc\u30f3\u63a8\u8ad6\u304a\u3088\u3073QA\u30c7\u30fc\u30bf\u30bb\u30c3\u30c8\u3067\u8a55\u4fa1\u3057\u305f\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2511.20857v1\">\u8ad6\u6587<\/a>\u00a0\u00a0<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2511.20857v1\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>\u00a0 \u00a0(Tue, 25 Nov 2025 21:08:07 GMT)<\/li>\n\n\n\n<li>\u300cThe benchmark covers both multi-turn goal-oriented environments and single-turn reasoning or problem-solving tasks, explicitly testing whether LLMs can accumulate knowledge and refine strategies during deployment, a process we term test-time evolution. We unify and implement over ten representative memory modules, including retrieval-based, workflow, and hierarchical memory systems, to study their adaptation behavior. To further examine experience reuse, we introduce ExpRAG, a simple retrieval-based baseline that leverages prior task experiences, and further develop ReMem, an advanced action\u2013think\u2013memory refine pipeline that tightly integrates reasoning, action, and memory updates for continual improvement.\u300d\u3068\u306e\u3053\u3068\u3002\u6bd4\u8f03\u304c\u96e3\u3057\u3044\u5206\u91ce\u3067\u3042\u308a\u3068\u3066\u3082\u3042\u308a\u304c\u305f\u3044\u30d9\u30f3\u30c1\u30de\u30fc\u30af\u3002\u30b7\u30f3\u30d7\u30eb\u306a\u6226\u7565\u304c\u597d\u30b9\u30b3\u30a2\u3092\u51fa\u3057\u3066\u3044\u308b\u70b9\u3082\u8208\u5473\u6df1\u3044\u3002\u3002\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":[244,517],"class_list":["post-7845","post","type-post","status-publish","format-standard","hentry","category-arxiv","tag-memory","tag-517"],"_links":{"self":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/7845","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=7845"}],"version-history":[{"count":1,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/7845\/revisions"}],"predecessor-version":[{"id":7846,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/7845\/revisions\/7846"}],"wp:attachment":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=7845"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=7845"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=7845"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}