{"id":9006,"date":"2026-06-10T08:05:10","date_gmt":"2026-06-10T00:05:10","guid":{"rendered":"https:\/\/infernews.com\/blog\/agents-39-last-exam\/"},"modified":"2026-06-10T08:05:10","modified_gmt":"2026-06-10T00:05:10","slug":"agents-39-last-exam","status":"publish","type":"post","link":"https:\/\/infernews.com\/blog\/agents-39-last-exam\/","title":{"rendered":"Agents&#8217; Last Exam \u8a55\u6e2c\u6846\u67b6\uff1a\u7576 AI \u4ee3\u7406\u8003\u5c08\u696d\u8a66"},"content":{"rendered":"<figure class=\"wp-block-image size-large\"><img decoding=\"async\" src=\"https:\/\/infernews.com\/blog\/wp-content\/uploads\/2026\/06\/teaser-d437a804e1cd.jpg\" alt=\"logo\"><\/figure>\n<p>\u9019\u500b\u958b\u6e90\u9805\u76ee\u662f\u4e00\u5957<strong>AI \u4ee3\u7406\u4eba\u8a55\u6e2c\u57fa\u6e96<\/strong>\uff0c\u5c08\u9580\u7528\u4f86\u91cf\u5ea6 Computer-use agents\uff08CUAs\uff09\u5728\u9577\u6642\u9593\u3001\u8de8\u6b65\u9a5f\u7684\u771f\u5be6\u5de5\u4f5c\u6d41\u7a0b\u4e2d\u8868\u73fe\u5982\u4f55\u3002\u5b83\u7684\u8a2d\u8a08\u76ee\u6a19\u662f\u8b93\u696d\u754c\u80fd\u7528\u540c\u4e00\u628a\u5c3a\uff0c\u53bb\u6bd4\u8f03 Claude Code\u3001Codex\u3001Openclaw \u7b49\u4e0d\u540c\u4ee3\u7406\u4eba\u5728\u771f\u5be6\u5834\u666f\u4e0b\u7684\u80fd\u529b\u5dee\u8ddd\uff0c\u800c\u4e0d\u518d\u505c\u7559\u5728\u7c21\u55ae\u7684\u554f\u7b54\u6e2c\u8a66\u3002<\/p>\n<p><strong>ALE<\/strong> \u8207\u574a\u9593\u5e38\u898b\u7684\u6392\u884c\u699c\u6700\u5927\u5206\u5225\uff0c\u5728\u65bc\u5b83\u628a\u4efb\u52d9\u5c0d\u9f4a\u5230\u7f8e\u570b O*NET \/ SOC 2018 \u8077\u696d\u5206\u985e\uff0c\u6db5\u84cb 55 \u500b\u975e\u5be6\u9ad4\u884c\u696d\uff0c\u4f8b\u5982\u52d5\u756b\u3001\u5de5\u7a0b\u88fd\u9020\u3001\u5efa\u7bc9\u8a2d\u8a08\u8207\u795e\u7d93\u5f71\u50cf\u5206\u6790\u7b49\u3002\u57fa\u6e96\u7684\u984c\u76ee\u4e26\u975e\u7531\u7814\u7a76\u54e1\u6191\u7a7a\u8a2d\u8a08\uff0c\u800c\u662f\u7531 300 \u591a\u4f4d\u696d\u754c\u5c08\u5bb6\u6536\u96c6\u4e26\u9a57\u8b49\uff0c\u56e0\u6b64\u6bcf\u4e00\u689d\u4efb\u52d9\u90fd\u8cbc\u8fd1\u771f\u5be6\u5de5\u4f5c\u5834\u666f\uff0c\u800c\u4e14\u914d\u6709\u96b1\u85cf\u53c3\u8003\u7b54\u6848\u8207\u78ba\u5b9a\u6027\u8a55\u5206\u6a5f\u5236\uff0c\u675c\u7d55\u6a21\u578b\u300c\u8d70\u65c1\u9580\u5de6\u9053\u300d\u8499\u6df7\u904e\u95dc\u3002<\/p>\n<p>\u5728\u6280\u8853\u67b6\u69cb\u4e0a\uff0c\u8a55\u6e2c\u6846\u67b6\u7531\u4e09\u90e8\u5206\u7d44\u6210\uff1a<strong>\u4ee3\u7406\u57f7\u884c\u74b0\u5883\uff08harness\uff09<\/strong> \u8ca0\u8cac\u9a45\u52d5\u5e95\u5c64\u57fa\u790e\u6a21\u578b\uff1b<strong>\u865b\u64ec\u6c99\u7bb1\uff08sandbox\uff09<\/strong> \u6a21\u64ec\u771f\u5be6\u4f5c\u696d\u7cfb\u7d71\u74b0\u5883\uff1b<strong>\u4efb\u52d9\u55ae\u5143\uff08task\uff09<\/strong> \u5247\u4ee5\u53ef\u57f7\u884c\u65b9\u5f0f\u63cf\u8ff0\u4e00\u9805\u5c08\u696d\u5de5\u4f5c\u3002\u76ee\u524d\u516c\u958b\u7684 150 \u689d\u4efb\u52d9\u5c6c\u65bc 1,500 \u984c\u4ee5\u4e0a\u8a9e\u6599\u5eab\u7684\u5b50\u96c6\uff0c\u6574\u500b\u9805\u76ee\u6b63\u5728\u671d 5,000 \u984c\u76ee\u6a19\u63a8\u9032\u3002\u8a55\u6e2c\u6642\uff0c\u6846\u67b6\u6703\u5728\u96f2\u7aef\u5efa\u7acb\u6c99\u7bb1\u3001\u8b93\u4ee3\u7406\u4eba\u5b8c\u6210\u4efb\u52d9\uff0c\u518d\u4ee5\u78ba\u5b9a\u6027\u8a55\u5206\u5668\u8a55\u5206\uff0c\u904e\u7a0b\u5168\u81ea\u52d5\u5316\u3002<\/p>\n<p>\u6e2c\u8a66\u524d\u9700\u8981\u5148\u5728 Google Cloud \u5b8c\u6210\u4e00\u6b21\u6027\u8a2d\u5b9a\uff08\u7d04 10 \u5206\u9418\uff0c\u73fe\u6709 $300 \u7f8e\u5143\u514d\u8cbb\u984d\u5ea6\u8db3\u4ee5\u61c9\u4ed8\uff09\uff0c\u4e4b\u5f8c\u5373\u53ef\u5728\u81ea\u5bb6\u6a21\u578b\u4e0a\u57f7\u884c hello-world \u4efb\u52d9\u4ee5\u719f\u6089\u6d41\u7a0b\uff0c\u9032\u4e00\u6b65\u63db\u6210\u6b63\u5f0f\u4efb\u52d9\u3002\u9019\u500b\u9805\u76ee\u9069\u5408 AI \u4ee3\u7406\u4eba\u958b\u767c\u8005\u3001\u4f01\u696d\u63a1\u8cfc\u5718\u968a\u3001\u5b78\u8853\u7814\u7a76\u8005\uff0c\u4ee5\u53ca\u4efb\u4f55\u60f3\u7528\u5ba2\u89c0\u5206\u6578\u6bd4\u8f03\u4e0d\u540c CUA \u8868\u73fe\u7684\u4eba\u3002<\/p>\n<ul>\n<li><strong>\u4ee3\u7406\u8a55\u6e2c\u57fa\u6e96<\/strong>\uff1a\u5c08\u70ba Computer-use agents \u8a2d\u8a08\u7684\u771f\u5be6\u5de5\u4f5c\u6d41\u7a0b\u6e2c\u8a66\u3002<\/li>\n<li><strong>\u884c\u696d\u8986\u84cb\u5ee3<\/strong>\uff1a\u5c0d\u9f4a O*NET \/ SOC 2018\uff0c\u6db5\u84cb 55 \u500b\u884c\u696d\uff0c\u516c\u958b 150 \u689d\u4efb\u52d9\u3002<\/li>\n<li><strong>\u7d50\u679c\u53ef\u9a57\u8b49<\/strong>\uff1a\u63a1\u7528\u96b1\u85cf\u53c3\u8003\u7b54\u6848\u8207\u78ba\u5b9a\u6027\u8a55\u5206\u5668\uff0c\u516c\u5e73\u4e14\u53ef\u91cd\u73fe\u3002<\/li>\n<li><strong>\u96f2\u7aef\u6c99\u7bb1\u74b0\u5883<\/strong>\uff1a\u5728 Google Cloud \u5167\u5efa\u7acb\u865b\u64ec\u6a5f\u91cd\u73fe\u771f\u5be6\u4f5c\u696d\u6d41\u7a0b\u3002<\/li>\n<li><strong>\u793e\u7fa4\u5171\u5275<\/strong>\uff1a\u7531\u67cf\u514b\u840a RDI \u8207 300 \u591a\u4f4d\u696d\u754c\u5c08\u5bb6\u5171\u540c\u7dad\u8b77\u3002<\/li>\n<\/ul>\n<p><strong>GitHub\uff1a<\/strong> <a href=\"https:\/\/github.com\/rdi-berkeley\/agents-last-exam\" rel=\"noopener noreferrer\">https:\/\/github.com\/rdi-berkeley\/agents-last-exam<\/a><\/p>\n<p><strong>\u9805\u76ee\uff1a<\/strong> <a href=\"https:\/\/agents-last-exam.org\/\" rel=\"noopener noreferrer\">https:\/\/agents-last-exam.org\/<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"<p>\u7531\u67cf\u514b\u840a RDI \u8207\u696d\u754c\u5c08\u5bb6\u5171\u540c\u6253\u9020\u7684\u958b\u6e90\u8a55\u6e2c\u9805\u76ee\uff0c\u7528 150 \u689d\u771f\u5be6\u5c08\u696d\u4efb\u52d9\u8003\u9a57 AI \u4ee3\u7406\u4eba\uff0c\u4e26\u4ee5\u53ef\u9a57\u8b49\u7d50\u679c\u91cf\u5316\u8868\u73fe\u3002<\/p>\n","protected":false},"author":8,"featured_media":9005,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"ai_generated_summary":"","footnotes":""},"categories":[133,116,197],"tags":[],"class_list":["post-9006","post","type-post","status-publish","format-standard","hentry","category-133","category-agentic","category-framework"],"_links":{"self":[{"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/posts\/9006","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/users\/8"}],"replies":[{"embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/comments?post=9006"}],"version-history":[{"count":0,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/posts\/9006\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/media\/9005"}],"wp:attachment":[{"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/media?parent=9006"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/categories?post=9006"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/tags?post=9006"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}