{"id":8334,"date":"2026-03-17T04:42:00","date_gmt":"2026-03-16T19:42:00","guid":{"rendered":"https:\/\/devneko.jp\/wordpress\/?p=8334"},"modified":"2026-03-14T16:46:42","modified_gmt":"2026-03-14T07:46:42","slug":"memo-memory-augmented-model-context-optimization-for-robust-multi-turn-multi-agent-llm-games","status":"publish","type":"post","link":"https:\/\/devneko.jp\/wordpress\/?p=8334","title":{"rendered":"MEMO: Memory-Augmented Model Context Optimization for Robust Multi-Turn Multi-Agent LLM Games"},"content":{"rendered":"\n<ul class=\"wp-block-list\">\n<li><strong>MEMO: Memory-Augmented Model Context Optimization for Robust Multi-Turn Multi-Agent LLM Games\u00a0<\/strong>[79.7]<br>\u30de\u30eb\u30c1\u30a8\u30fc\u30b8\u30a7\u30f3\u30c8\u30b2\u30fc\u30e0\u306e\u8a55\u4fa1\u306f\u3001\u3057\u3070\u3057\u3070\u5b9f\u8cea\u7684\u306a\u30e9\u30f3\u30fb\u30c4\u30fc\u30fb\u30e9\u30f3\u306e\u3070\u3089\u3064\u304d\u3092\u793a\u3059\u3002 \u56de\u8ee2\u3059\u308b\u5c0f\u3055\u306a\u504f\u5dee\u306f\u3001\u591a\u30a8\u30fc\u30b8\u30a7\u30f3\u30c8\u30ab\u30c3\u30d7\u30ea\u30f3\u30b0\u306b\u3088\u3063\u3066\u5897\u5e45\u3055\u308c\u308b\u3002 \u81ea\u5df1\u518d\u751f\u30d5\u30ec\u30fc\u30e0\u30ef\u30fc\u30af\u3067\u3042\u308bMEMO(Memory-augmented MOdel context optimization)\u3092\u7528\u3044\u3066,\u4e0d\u5b89\u5b9a\u6027\u3068\u30a2\u30f3\u30c0\u30fc\u30d1\u30d5\u30a9\u30fc\u30de\u30f3\u30b9\u306e\u4e21\u9762\u306b\u5bfe\u51e6\u3059\u308b\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2603.09022v1\">\u8ad6\u6587<\/a>\u00a0\u00a0<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2603.09022v1\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>\u00a0 \u00a0(Mon, 09 Mar 2026 23:36:32 GMT)<\/li>\n\n\n\n<li>\u300cWe therefore propose MEMO (Memory-augmented MOdel context optimization), a self- play framework that optimizes inference-time context without updating model weights. \u300d\u300cThe central finding is that exploration alone yields only modest gains; persistent memory is what transforms context optimization from a memoryless search into a cumulative learning process.\u300d\u3068\u30b2\u30fc\u30e0\u306b\u304a\u3051\u308b\u8a18\u61b6\u306e\u6709\u52b9\u6027\u3092\u4e3b\u5f35\u3002<\/li>\n\n\n\n<li>\u30d7\u30ed\u30b8\u30a7\u30af\u30c8\u30b5\u30a4\u30c8\u306f<a href=\"https:\/\/yunfeixie233.github.io\/MEMO\/\">MEMO: Memory-Augmented Model Context Optimization for Robust Multi-Turn Multi-Agent LLM Games<\/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":[244],"class_list":["post-8334","post","type-post","status-publish","format-standard","hentry","category-arxiv","tag-memory"],"_links":{"self":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/8334","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=8334"}],"version-history":[{"count":1,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/8334\/revisions"}],"predecessor-version":[{"id":8335,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/8334\/revisions\/8335"}],"wp:attachment":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=8334"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=8334"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=8334"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}