
{"id":9834,"date":"2026-07-03T21:18:50","date_gmt":"2026-07-03T13:18:50","guid":{"rendered":"https:\/\/infernews.com\/blog\/agenticdatabench\/"},"modified":"2026-07-03T21:18:50","modified_gmt":"2026-07-03T13:18:50","slug":"agenticdatabench","status":"publish","type":"post","link":"https:\/\/infernews.com\/blog\/agenticdatabench\/","title":{"rendered":"AgenticDataBench\uff1a\u6578\u64da\u4ee3\u7406\u57fa\u6e96\u9ede\u6a23\u7747"},"content":{"rendered":"\n<figure class=\"wp-block-image size-large\"><img decoding=\"async\" src=\"https:\/\/infernews.com\/blog\/wp-content\/uploads\/2026\/07\/example.png\" alt=\"example\"><\/figure>\n\n\n\n<p class=\"wp-block-paragraph\">AgenticDataBench \u662f\u4e00\u500b\u7528\u4f86\u8a55\u6e2c data agents \u7684 benchmark\uff0c\u800c\u5514\u4fc2\u76f4\u63a5\u5e6b\u4eba\u505a\u5206\u6790\u7684\u6a21\u578b\u6216\u61c9\u7528\u3002\u5b83\u8981\u89e3\u6c7a\u7684\u662f\uff1aLLM-based data agents \u80fd\u5426\u7a69\u5b9a\u5b8c\u6210 data science workflow\uff0c\u4e26\u4e14\u7528\u53ef\u6bd4\u8f03\u3001\u53ef\u91cd\u73fe\u7684\u65b9\u5f0f\u91cf\u5ea6\u8868\u73fe\u3002<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">\u73fe\u6709\u505a\u6cd5\u591a\u6578\u53ea\u7528\u96f6\u6563\u4efb\u52d9\u3001\u55ae\u4e00\u8cc7\u6599\u96c6\uff0c\u6216\u8005\u53ea\u770b\u6700\u7d42\u7b54\u6848\uff0c\u8f03\u96e3\u77e5\u9053\u4ee3\u7406\u7a76\u7adf\u5361\u5728\u54ea\u500b\u6b65\u9a5f\u3002\u9019\u500b\u9805\u76ee\u6539\u7528 <strong>344 \u500b\u4efb\u52d9\u300115 \u500b\u9818\u57df<\/strong>\uff0c\u518d\u914d\u5408\u7d30\u7dfb\u7684 skill labels \u540c ground-truth\uff0c\u5c07\u554f\u984c\u62c6\u6210\u53ef\u91cd\u7528\u7684 data science skills\uff0c\u4f8b\u5982\u7f3a\u5931\u503c\u8655\u7406\u4e00\u985e\u64cd\u4f5c\u6a21\u5f0f\uff0c\u4ee4\u8a55\u6e2c\u5514\u53ea\u5f97\u7e3d\u5206\uff0c\u4ef2\u53ef\u4ee5\u898b\u5230\u6280\u80fd\u5c64\u9762\u7684\u5f37\u5f31\u3002<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">\u90e8\u7f72\u540c\u7406\u89e3\u65b9\u5f0f\u90fd\u5e7e\u76f4\u63a5\uff1a\u8cc7\u6599\u96c6\u53ef\u7531 HuggingFace \u4e0b\u8f09\u5f8c\u653e\u5165 <code>testbed\/datasets\/<\/code>\uff0c\u4efb\u52d9\u3001gold \u6a19\u8a3b\u540c\u7d50\u679c\u76ee\u9304\u5df2\u7d93\u5206\u958b\uff0c\u53e6\u5916\u4fdd\u7559\u5497 98 \u500b private test tasks \u7dad\u6301 leaderboard \u7684\u53ef\u4fe1\u5ea6\u3002README \u4ea6\u63d0\u5230\u9700\u8981\u8a2d\u5b9a API keys\uff0c\u53cd\u6620\u5b83\u4e3b\u8981\u4fc2\u4e00\u500b\u958b\u653e\u6e2c\u8a66\u53f0\uff0c\u65b9\u4fbf\u7528\u4e0d\u540c agent harness \u8dd1\u540c\u4e00\u6279\u4efb\u52d9\uff0c\u800c\u5514\u4fc2\u55ae\u6a5f\u5373\u958b\u5373\u7528\u7684\u7d42\u7aef\u5de5\u5177\u3002<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">\u540c\u985e benchmark \u76f8\u6bd4\uff0c\u5b83\u7684\u53d6\u5411\u5514\u4fc2\u8ffd\u6c42\u6700\u5c11\u984c\u76ee\u4e0b\u7684\u5feb\u901f\u6392\u884c\uff0c\u800c\u4fc2\u5f37\u8abf<strong>\u771f\u5be6\u6027\u3001\u6280\u80fd\u8986\u84cb\u7387\u540c\u5197\u9918\u63a7\u5236<\/strong>\u3002\u9805\u76ee\u4e00\u65b9\u9762\u6536\u9304\u771f\u5be6 B2B fintech use cases\uff0c\u53e6\u4e00\u65b9\u9762\u7528 skill-aligned hierarchical clustering \u540c\u7cfb\u7d71\u5316\u751f\u6210\u6d41\u7a0b\u88dc\u8db3\u7f3a\u5c11\u771f\u5be6\u4efb\u52d9\u7684\u9818\u57df\uff0c\u9019\u7a2e\u505a\u6cd5\u7684\u4ee3\u50f9\u662f\u5efa\u7f6e\u8207\u7dad\u8b77\u8f03\u91cd\uff0c\u4f46\u63db\u4f86\u66f4\u5b8c\u6574\u7684\u6bd4\u8f03\u57fa\u7dda\u3002<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>\u8986\u84cb 15 \u500b\u9818\u57df\uff0c\u5305\u542b\u771f\u5be6 B2B fintech \u4efb\u52d9<\/li>\n<li>\u63d0\u4f9b tasks\u3001ground-truth\u3001skills \u540c results \u7d50\u69cb\u5316\u5167\u5bb9<\/li>\n<li>\u652f\u63f4\u6bd4\u8f03\u4e0d\u540c agent harness\uff0c\u5982 Smolagents\u3001DA-Agent\u3001Claude Code\u3001CodeX<\/li>\n<li>\u5df2\u5217\u51fa Qwen3.5-397B-A17B\u3001Kimi-K2.5\u3001Claude Sonnet 4.6 \u7684\u521d\u6b65\u5be6\u9a57<\/li>\n<\/ul>\n\n\n\n<p class=\"wp-block-paragraph\">\u9019\u500b\u9805\u76ee\u6700\u9069\u5408\u505a data agent \u7814\u767c\u3001\u6a21\u578b\u9078\u578b\u540c\u5167\u90e8\u9a57\u8b49\u7684\u5718\u968a\uff0c\u4e5f\u9069\u5408\u7814\u7a76\u4eba\u54e1\u7528\u4f86\u6aa2\u67e5\u4ee3\u7406\u5728\u54ea\u985e data skills \u5931\u5206\u3002\u6027\u80fd\u8cc7\u8a0a\u76ee\u524d\u4ee5 leaderboard \u7d50\u679c\u70ba\u4e3b\uff0c\u91cd\u9ede\u4e0d\u53ea\u662f accuracy\uff0c\u4ef2\u5305\u62ec skill-level insight\uff1b\u76f8\u95dc\u6a21\u578b\u81f3\u5c11\u5305\u62ec Qwen3.5-397B-A17B\u3001Kimi-K2.5 \u540c Claude Sonnet 4.6\u3002<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/agenticdatabench.github.io\/#\/\" rel=\"noopener noreferrer\" target=\"_blank\"><strong>\u9805\u76ee\u4e3b\u9801<\/strong><\/a> \u00b7 <a href=\"https:\/\/github.com\/AgenticDataBench\/AgenticDataBench\" rel=\"noopener noreferrer\" target=\"_blank\"><strong>GitHub<\/strong><\/a> \u00b7 <a href=\"https:\/\/arxiv.org\/pdf\/2607.01647\" rel=\"noopener noreferrer\" target=\"_blank\"><strong>Paper<\/strong><\/a><\/p>\n","protected":false},"excerpt":{"rendered":"<p>\u9019\u662f\u4e00\u500b\u8a55\u6e2c data agents \u7684 benchmark\u3002\u5b83\u7528\u771f\u5be6\u4efb\u52d9\u540c\u6280\u80fd\u6a19\u8a3b\uff0c\u6aa2\u67e5\u6a21\u578b\u662f\u5426\u771f\u4fc2\u8b58\u505a\u6578\u64da\u5de5\u4f5c\u3002<\/p>\n","protected":false},"author":8,"featured_media":9833,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"ai_generated_summary":"","footnotes":""},"categories":[133,185,116,137,191,196,199,201],"tags":[],"class_list":["post-9834","post","type-post","status-publish","format-standard","hentry","category-133","category-qwen","category-agentic","category-api","category-anthropic","category-196","category-dataset-","category-skill-"],"_links":{"self":[{"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/posts\/9834","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=9834"}],"version-history":[{"count":0,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/posts\/9834\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/media\/9833"}],"wp:attachment":[{"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/media?parent=9834"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/categories?post=9834"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/tags?post=9834"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}