{"id":5144,"date":"2024-07-26T06:30:00","date_gmt":"2024-07-25T21:30:00","guid":{"rendered":"https:\/\/devneko.jp\/wordpress\/?p=5144"},"modified":"2024-07-26T06:30:00","modified_gmt":"2024-07-25T21:30:00","slug":"gta-a-benchmark-for-general-tool-agents","status":"publish","type":"post","link":"https:\/\/devneko.jp\/wordpress\/?p=5144","title":{"rendered":"GTA: A Benchmark for General Tool Agents\u00a0"},"content":{"rendered":"\n<ul class=\"wp-block-list\">\n<li><strong>GTA: A Benchmark for General Tool Agents\u00a0<\/strong>[32.4]<br>229\u500b\u306e\u5b9f\u4e16\u754c\u306e\u30bf\u30b9\u30af\u3068\u5b9f\u884c\u53ef\u80fd\u306a\u30c4\u30fc\u30eb\u30c1\u30a7\u30fc\u30f3\u3092\u8a2d\u8a08\u3057\u3001\u4e3b\u8981\u306a\u5927\u8a00\u8a9e\u30e2\u30c7\u30eb(LLM)\u3092\u8a55\u4fa1\u3059\u308b\u3002 GPT-4 \u306f\u30bf\u30b9\u30af\u306e 50% \u4ee5\u4e0b\u3067\u3042\u308a,\u307b\u3068\u3093\u3069\u306e LLM \u306f 25% \u4ee5\u4e0b\u3067\u3042\u308b\u3002 \u3053\u306e\u8a55\u4fa1\u306f\u3001\u73fe\u5b9f\u30b7\u30ca\u30ea\u30aa\u306b\u304a\u3051\u308b\u73fe\u5728\u306eLLM\u306e\u30c4\u30fc\u30eb\u5229\u7528\u80fd\u529b\u306e\u30dc\u30c8\u30eb\u30cd\u30c3\u30af\u3092\u660e\u3089\u304b\u306b\u3057\u3001\u6c4e\u7528\u30c4\u30fc\u30eb\u30a8\u30fc\u30b8\u30a7\u30f3\u30c8\u3092\u524d\u9032\u3055\u305b\u308b\u305f\u3081\u306e\u4eca\u5f8c\u306e\u65b9\u5411\u6027\u3092\u63d0\u4f9b\u3059\u308b\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2407.08713v1\">\u8ad6\u6587<\/a>\u00a0\u00a0<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2407.08713v1\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>\u00a0 \u00a0(Thu, 11 Jul 2024 17:50:09 GMT)<\/li>\n\n\n\n<li>\u30ea\u30a2\u30eb\u306a\u30b7\u30ca\u30ea\u30aa\u3067\u30c4\u30fc\u30eb\u3092\u5229\u7528\u3057\u3066\u554f\u984c\u3092\u89e3\u304f\u30a8\u30fc\u30b8\u30a7\u30f3\u30c8\u306e\u80fd\u529b\u3092\u8a55\u4fa1\u3059\u308b\u30d9\u30f3\u30c1\u30de\u30fc\u30af\u3002\u60f3\u5b9a\u3055\u308c\u3066\u3044\u308b\u30c4\u30fc\u30eb\u306fOCR\u3084DrawBox\u3001Calculator\u3001TextToImage\u306a\u306914\u7a2e\u985e\u3002\u300cOur findings show that realworld user queries are challenging for existing LLMs, with GPT-4 completing less than 50% of the tasks and most LLMs achieving below 25%.\u300d\u3068\u306e\u3053\u3068\u3002<\/li>\n\n\n\n<li>\u30ea\u30dd\u30b8\u30c8\u30ea\u306f<a href=\"https:\/\/github.com\/open-compass\/GTA\">GitHub &#8211; open-compass\/GTA: Official repository for paper &#8220;GTA: A Benchmark for General Tool Agents&#8221;<\/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":[42,517],"class_list":["post-5144","post","type-post","status-publish","format-standard","hentry","category-arxiv","tag-autonomous-agent","tag-517"],"_links":{"self":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/5144","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=5144"}],"version-history":[{"count":0,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/5144\/revisions"}],"wp:attachment":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=5144"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=5144"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=5144"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}