{"id":39649,"date":"2025-03-26T05:02:22","date_gmt":"2025-03-25T21:02:22","guid":{"rendered":"https:\/\/17aitech.com\/?p=39649"},"modified":"2025-03-26T05:02:22","modified_gmt":"2025-03-25T21:02:22","slug":"%e6%a3%8b%e7%9b%98%e5%8f%98%e6%88%98%e5%9c%ba%ef%bc%8c%e5%a4%a7%e6%a8%a1%e5%9e%8b%e5%8d%b4%e5%91%86%e4%ba%86%ef%bc%9f%e6%99%ae%e6%9e%97%e6%96%af%e9%a1%bf%e3%80%81ut-austin%e6%96%b0%e5%9f%ba%e5%87%86sp","status":"publish","type":"post","link":"https:\/\/17aitech.com\/?p=39649","title":{"rendered":"\u68cb\u76d8\u53d8\u6218\u573a\uff0c\u5927\u6a21\u578b\u5374\u5446\u4e86\uff1f\u666e\u6797\u65af\u987f\u3001UT Austin\u65b0\u57fa\u51c6SPIN-Bench\u66ddAI\u667a\u5546\u74f6\u9888"},"content":{"rendered":"<p>\u6587\u7ae0\u6765\u6e90\u4e8e\u4e92\u8054\u7f51:<a href=\"https:\/\/www.jiqizhixin.com\/articles\/2025-03-25-3\" target=\"_blank\">\u68cb\u76d8\u53d8\u6218\u573a\uff0c\u5927\u6a21\u578b\u5374\u5446\u4e86\uff1f\u666e\u6797\u65af\u987f\u3001UT Austin\u65b0\u57fa\u51c6SPIN-Bench\u66ddAI\u667a\u5546\u74f6\u9888<\/a><\/p>\n<section><a href=\"https:\/\/17aitech.com\/wp-content\/uploads\/2025\/03\/frc-84fab4aaea33ecaf86bc8a6a050a95b3.png\" data-fancybox=\"images\" data-fancybox=\"gallery\"><img decoding=\"async\" src=\"https:\/\/17aitech.com\/wp-content\/uploads\/2025\/03\/frc-84fab4aaea33ecaf86bc8a6a050a95b3.png\"><\/a><\/section>\n<section>\u5f53\u68cb\u76d8\u53d8\u6210\u6218\u573a\uff0c\u5f53\u76df\u53cb\u6697\u85cf\u5fc3\u673a\uff0c\u5f53\u8c08\u5224\u9700\u8981\u4e09\u5341\u516d\u8ba1\uff0cAI \u7684\u667a\u5546\u4ee4\u4eba\u53f9\u606f\uff01<\/section>\n<section><\/section>\n<section>\u8fd1\u65e5\uff0c\u6765\u81ea\u666e\u6797\u65af\u987f\u548c\u5fb7\u5dde\u5927\u5b66\u5965\u65af\u4e01\u5206\u6821<strong>\u6700\u65b0\u8bc4\u6d4b\u57fa\u51c6 SPIN-Bench\uff0c\u7528\u4e00\u5957 &#8220;\u7ec4\u5408\u62f3&#8221; \u66b4\u51fb\u4e86\u5927\u6a21\u578b\u7684\u8f6f\u808b\u3002<\/strong>\u7814\u7a76\u663e\u793a\uff0c<strong>\u5373\u4fbf\u662f o1\u3001o3-mini\u3001DeepSeek R1\u3001GPT-4o\u3001Claude 3.5 \u7b49\u9876\u5c16\u5927\u6a21\u578b\uff0c\u5728\u6d89\u53ca\u6218\u7565\u89c4\u5212\u548c\u793e\u4f1a\u63a8\u7406\u7684\u590d\u6742\u4efb\u52a1\u4e2d\u96c6\u4f53 &#8220;\u81ea\u95ed&#8221;\u3002<\/strong><\/section>\n<section><a href=\"https:\/\/17aitech.com\/wp-content\/uploads\/2025\/03\/frc-6a81a03b89a4cb11b8fb4910c62cd319.png\" data-fancybox=\"images\" data-fancybox=\"gallery\"><img decoding=\"async\" src=\"https:\/\/17aitech.com\/wp-content\/uploads\/2025\/03\/frc-6a81a03b89a4cb11b8fb4910c62cd319.png\"><\/a><\/section>\n<ul>\n<li>\n<section>\u8bba\u6587\u6807\u9898\uff1aSPIN-Bench: How Well Do LLMs Plan Strategically and Reason Socially?<\/section>\n<\/li>\n<li>\n<section>\u8bba\u6587\u94fe\u63a5\uff1ahttps:\/\/arxiv.org\/pdf\/2503.12349<\/section>\n<\/li>\n<li>\n<section>\u9879\u76ee\u4e3b\u9875\uff1ahttps:\/\/spinbench.github.io<\/section>\n<\/li>\n<\/ul>\n<section><\/section>\n<section>\u5728\u8fc7\u53bb\u7684\u4e00\u5e74\u91cc\uff0c\u5927\u8bed\u8a00\u6a21\u578b\uff08LLM\uff09\u5c55\u73b0\u4e86\u4ee4\u4eba\u60ca\u53f9\u7684 &#8220;\u6587\u672c\u751f\u6210&#8221; \u548c &#8220;\u667a\u80fd\u4ee3\u7406&#8221; \u80fd\u529b\u3002\u8bb8\u591a\u793e\u533a\u7528\u6237\u5df2\u7ecf\u89c1\u5230\u4e86\u5404\u5927\u6a21\u578b\u7684\u300c\u767e\u82b1\u9f50\u653e\u300d\uff1a\u4ece\u9ad8\u6548\u7684\u7b80\u5355\u95ee\u7b54\u5230\u591a\u8f6e\u5bf9\u8bdd\u63a8\u7406\u3002<\/section>\n<section><\/section>\n<section>\u7136\u800c\uff0c\u5f53\u6211\u4eec\u8c08\u5230\u771f\u6b63\u590d\u6742\u7684 &#8220;\u601d\u8003&#8221; \u573a\u666f \u2014\u2014 \u8b6c\u5982\u9700\u8981\u5728\u4e00\u4e2a\u5145\u6ee1\u5176\u4ed6 &#8220;\u73a9\u5bb6&#8221;\uff08\u4eba\u6216\u667a\u80fd\u4f53\uff09\u7684\u4e0d\u786e\u5b9a\u73af\u5883\u4e2d\u505a\u51fa\u957f\u7a0b\u7b56\u7565\u89c4\u5212\u3001\u4e34\u573a\u8c08\u5224\u5408\u4f5c\u751a\u81f3 &#8220;\u80cc\u540e\u4f7f\u574f&#8221; \u65f6\uff0c\u5f53\u4e0b\u7684 LLM \u662f\u5426\u8fd8\u80fd\u7ad9\u7a33\u811a\u8ddf\uff1f<\/section>\n<section><\/section>\n<section>\u4f20\u7edf AI \u6d4b\u8bd5\u603b\u8ba9\u5927\u6a21\u578b\u505a &#8220;\u4e56\u5b66\u751f&#8221;\uff1a\u89e3\u6570\u5b66\u9898\u3001\u5199\u4ee3\u7801\u3001\u80cc\u767e\u79d1&#8230;&#8230; \u4f46\u5728\u771f\u5b9e\u4e16\u754c\u4e2d\uff0c\u4eba\u7c7b\u66f4\u5e38\u7528\u7684\u667a\u80fd\u662f\u52a8\u6001\u535a\u5f08\u4e2d\u7684\u8c0b\u7565\u548c\u590d\u6742\u793e\u4ea4\u4e2d\u7684\u6d1e\u5bdf\u3002<\/section>\n<section><\/section>\n<section>\u4e3a\u89e3\u7b54\u8fd9\u4e00\u95ee\u9898\uff0c\u4f5c\u8005\u63a8\u51fa\u4e86\u5168\u65b0\u7684\u591a\u57df\u8bc4\u4f30\u6846\u67b6 SPIN-Bench\uff08Strategic Planning, Interaction, and Negotiation\uff09\uff0c<strong>\u5c06\u5355\u4eba\u89c4\u5212\u3001\u5408\u4f5c\u6e38\u620f\u3001\u5bf9\u6297\u535a\u5f08\u548c\u591a\u65b9\u8c08\u5224\u7edf\u4e00\u5230\u4e00\u4e2a\u6d4b\u8bd5\u6846\u67b6\u4e2d\uff0c\u5e76\u7cfb\u7edf\u5316\u5730\u6269\u5927\u73af\u5883\u89c4\u6a21\u548c\u590d\u6742\u5ea6<\/strong>\uff0c\u65e8\u5728\u5168\u9762\u523b\u753b LLM \u5728\u6218\u7565\u89c4\u5212\u4e0e\u793e\u4ea4\u63a8\u7406\u65b9\u9762\u7684 &#8220;\u77ed\u677f&#8221; \u4e0e\u6f5c\u529b\u3002<\/section>\n<section><strong><br \/><\/strong><\/section>\n<section><strong>SPIN-Bench: \u4e00\u89c8\u591a\u7ef4\u5ea6\u8bc4\u4f30\u4f53\u7cfb<\/strong><\/section>\n<section><a href=\"https:\/\/17aitech.com\/wp-content\/uploads\/2025\/03\/frc-3deffc3b82f76e1fff28b72367100098.png\" data-fancybox=\"images\" data-fancybox=\"gallery\"><img decoding=\"async\" src=\"https:\/\/17aitech.com\/wp-content\/uploads\/2025\/03\/frc-3deffc3b82f76e1fff28b72367100098.png\"><\/a><\/section>\n<section>SPIN-Bench \u6846\u67b6\u5305\u542b\u4e24\u4e2a\u6838\u5fc3\u7ec4\u6210\u90e8\u5206\uff1a<\/section>\n<section><\/section>\n<section>1. \u6e38\u620f\u4ee3\u7406\uff0c\u5305\u62ec LLM \u53ca\u5176\u81ea\u9002\u5e94\u63d0\u793a\uff1b<\/section>\n<section>2. \u73af\u5883\u548c\u8bc4\u4f30\u5b50\u7cfb\u7edf\uff0c\u7528\u4e8e\u7ba1\u7406\u6e38\u620f\u903b\u8f91\u3001\u8ddf\u8e2a\u4e92\u52a8\u548c\u91cf\u5316\u6027\u80fd\u3002<\/section>\n<section><\/section>\n<section>\u4f20\u7edf\u7684\u89c4\u5212\u8bc4\u6d4b\u5927\u591a\u5728\u5355\u4eba\u3001\u53ef\u5b8c\u5168\u89c2\u5bdf\u7684\u73af\u5883\u4e2d\u8fdb\u884c\uff0c\u65e0\u6cd5\u5145\u5206\u53cd\u6620\u73b0\u5b9e\u4e2d\u56e2\u961f\u51b3\u7b56\u7684\u590d\u6742\u5ea6\u3002\u800c SPIN-Bench \u8bd5\u56fe\u901a\u8fc7\u5f62\u5f0f\u5316\u4efb\u52a1\u4e0e\u591a\u4eba\u573a\u666f\u76f8\u7ed3\u5408\uff0c\u628a\u73b0\u5b9e\u4e2d\u9700\u8981\u7684 &#8220;\u540c\u4f34\u5408\u4f5c&#8221;&#8221; \u8c08\u5224\u535a\u5f08 &#8221; \u7b49\u5173\u952e\u6280\u80fd\u4e00\u5e76\u7eb3\u5165\uff0c\u4ee5\u5e2e\u52a9\u627e\u5230 LLM \u5728\u771f\u5b9e\u5e94\u7528\u843d\u5730\u65f6\u53ef\u80fd\u9762\u4e34\u7684\u95ee\u9898\u3002<\/section>\n<section><\/section>\n<section>SPIN-Bench \u8ba9 LLM \u9762\u5bf9\u4ece &#8220;\u6700\u57fa\u7840&#8221; \u7684\u5355\u667a\u80fd\u4f53\u89c4\u5212\uff0c\u5230 &#8220;\u6700\u590d\u6742&#8221; \u7684\u591a\u667a\u80fd\u4f53\u6df7\u5408\u5bf9\u6297\u4e0e\u5408\u4f5c\uff0c\u9010\u6b65\u5347\u7ea7\u96be\u5ea6\u3002\u6587\u7ae0\u5e76\u4e0d\u4ec5\u4ec5\u7edf\u8ba1 &#8220;\u6700\u7ec8\u662f\u5426\u83b7\u80dc&#8221; \u6216\u8005 &#8220;\u662f\u5426\u8fbe\u6210\u76ee\u6807&#8221;\uff0c\u8fd8\u989d\u5916\u8bbe\u7f6e\u4e86\u82e5\u5e72\u7ec6\u9897\u7c92\u5ea6\u7684\u6307\u6807\uff0c\u7528\u6765\u91cf\u5316\u6a21\u578b\u7684\u51b3\u7b56\u51c6\u786e\u6027\u3001\u534f\u4f5c\u6709\u6548\u6027\u4ee5\u53ca\u5728\u793e\u4ea4\u573a\u666f\u4e0b\u7684\u8bdd\u672f\u4e0e\u6267\u884c\u5339\u914d\u5ea6\u3002<\/section>\n<section><\/section>\n<section>\u5177\u4f53\u800c\u8a00\uff0c\u8be5\u6587\u4e3b\u8981\u805a\u7126\u4e09\u4e2a\u5c42\u6b21\uff1a<\/section>\n<section><\/section>\n<ul>\n<li>\n<section><strong>\u7ecf\u5178\u89c4\u5212\uff08PDDL\uff09- \u6d4b\u8bd5 &#8220;\u5355\u667a\u80fd\u4f53\u601d\u7ef4&#8221; \u6781\u9650<\/strong><\/section>\n<\/li>\n<\/ul>\n<section><\/section>\n<section>1\u3001\u5355\u667a\u80fd\u4f53\uff0c\u786e\u5b9a\u6027\u73af\u5883\uff0c\u591a\u6b65\u52a8\u4f5c\u89c4\u5212\uff0c\u901a\u8fc7\u5c42\u5c42\u9012\u8fdb\u7684\u96be\u5ea6\uff0c\u5206\u6790\u6a21\u578b\u7684\u9519\u8bef\u539f\u56e0\u3002<\/section>\n<section><\/section>\n<section>2\u3001\u6db5\u76d6 21 \u4e2a\u9886\u57df\uff08\u5171 1,280 \u4e2a\u4efb\u52a1\uff09\uff0c\u5305\u542b elevator\u3001grid\u3001floortile \u7b49\u591a\u4e2a\u5e38\u89c1\u5b50\u4efb\u52a1\uff0c\u8003\u5bdf\u70b9\u6d89\u53ca\u72b6\u6001\u7a7a\u95f4\u7684\u9010\u6b65\u63d0\u5347\u548c\u9010\u6e10\u590d\u6742\u7684\u7ea6\u675f\u6761\u4ef6\u3002<\/section>\n<section><\/section>\n<section>3\u3001\u5728\u7ecf\u5178\u89c4\u5212\u95ee\u9898\u4e2d\uff0c\u9898\u76ee\u901a\u5e38\u4f1a\u660e\u786e\u7ed9\u51fa\u521d\u59cb\u72b6\u6001\u3001\u53ef\u6267\u884c\u7684\u52a8\u4f5c\u96c6\u4ee5\u53ca\u5e0c\u671b\u8fbe\u5230\u7684\u76ee\u6807\u72b6\u6001\u3002Agent \u7684\u4efb\u52a1\u5219\u662f\u5229\u7528\u8fd9\u4e9b\u5df2\u77e5\u4fe1\u606f\uff0c\u89c4\u5212\u5e76\u751f\u6210\u4ece\u521d\u59cb\u72b6\u6001\u5230\u76ee\u6807\u72b6\u6001\u7684\u52a8\u4f5c\u5e8f\u5217\uff08trajectory\uff09\u3002<\/section>\n<section><\/section>\n<section>\u4f8b\u5982\uff0c\u5728\u4e00\u4e2a\u7535\u68af\u63a7\u5236\u95ee\u9898\u4e2d\uff0cAgent \u53ef\u4ee5\u6267\u884c\u7535\u68af\u7684\u4e0a\u4e0b\u79fb\u52a8\u548c\u5f00\u5173\u95e8\u7b49\u52a8\u4f5c\uff0c\u5b83\u9700\u8981\u901a\u8fc7\u5408\u7406\u89c4\u5212\uff0c\u5728\u6700\u5c11\u7684\u6b65\u9aa4\u5185\uff0c\u5c06\u6240\u6709\u4e58\u5ba2\u51c6\u786e\u3001\u9ad8\u6548\u5730\u8fd0\u9001\u5230\u4ed6\u4eec\u5404\u81ea\u5bf9\u5e94\u7684\u76ee\u6807\u697c\u5c42\u3002\u5728\u8fd9\u4e2a\u4f8b\u5b50\u4e2d o1 \u628a\u6700\u540e\u4e00\u4f4d\u4e58\u5ba2\uff08p4\uff09\u9001\u9519\u4e86\u697c\u5c42\uff0c\u8bf4\u660e LLM \u4ecd\u6709\u63d0\u5347\u7a7a\u95f4\u3002<\/section>\n<section><a href=\"https:\/\/17aitech.com\/wp-content\/uploads\/2025\/03\/frc-f3cec67f6518046ffa5130e6fb9294a4.png\" data-fancybox=\"images\" data-fancybox=\"gallery\"><img decoding=\"async\" src=\"https:\/\/17aitech.com\/wp-content\/uploads\/2025\/03\/frc-f3cec67f6518046ffa5130e6fb9294a4.png\"><\/a><\/section>\n<ul>\n<li>\n<section><strong>\u591a\u667a\u80fd\u4f53\u5bf9\u6297\u6216\u5408\u4f5c &#8211; \u9010\u6b65\u5347\u7ea7\u7684\u8111\u529b\u7ede\u6740<\/strong><\/section>\n<\/li>\n<\/ul>\n<section><\/section>\n<section>1. \u5bf9\u6297\u573a\u666f\uff08\u4e95\u5b57\u68cb\uff0c\u56db\u5b50\u68cb\uff0c\u56fd\u9645\u8c61\u68cb\uff09\uff1a\u6587\u7ae0\u5206\u522b\u5bf9\u6bd4 LLM \u7684\u843d\u5b50\u884c\u52a8\u4e0e\u4e13\u4e1a\u535a\u5f08\u5f15\u64ce\u3001\u542f\u53d1\u5f0f\u641c\u7d22\u7b97\u6cd5\u7684\u5bf9\u6218\u5e73\u5c40\u6982\u7387\u4e0e\u9009\u6b65\u5dee\u5f02\uff0c\u52a8\u4f5c\u662f\u5426\u7b26\u5408\u6700\u4f18\u7b56\u7565\uff0c\u8bc4\u4f30\u5176\u5728\u6218\u672f\u548c\u6218\u7565\u5c42\u9762\u7684\u6df1\u5ea6\u3002<\/section>\n<section><\/section>\n<section>2. \u5408\u4f5c\u573a\u666f\uff08Hanabi\uff09\uff1a\u8003\u6838\u6a21\u578b\u9762\u5bf9\u4e0d\u5b8c\u5168\u4fe1\u606f\u65f6\uff0c\u662f\u5426\u80fd\u591f\u901a\u8fc7\u6c9f\u901a\u9690\u542b\u4fe1\u606f\u3001\u63a8\u6d4b\u961f\u53cb\u624b\u724c\uff0c\u5b9e\u73b0\u56e2\u4f53\u534f\u4f5c\u3002<\/section>\n<section><\/section>\n<ul>\n<li>\n<section><strong>\u6218\u7565\u6e38\u620f\u4e0e\u8c08\u5224 &#8211; \u4e03\u56fd\u6df7\u6218 + \u5b9e\u65f6\u8c08\u5224\uff0c\u4e0a\u6f14 AI \u7248\u300a\u6743\u529b\u7684\u6e38\u620f\u300b<\/strong><\/section>\n<\/li>\n<\/ul>\n<section><\/section>\n<section>Diplomacy \u662f\u4e00\u6b3e\u878d\u5408\u8054\u76df\u3001\u8c08\u5224\u3001\u80cc\u53db\u4e0e\u5408\u4f5c\u7684\u7b56\u7565\u7c7b\u684c\u6e38\u3002\u73a9\u5bb6\u4e4b\u95f4\u9700\u8981\u76f8\u4e92\u901a\u4fe1\u3001\u7ed3\u76df\u6216\u6b3a\u9a97\uff0c\u6700\u7ec8\u540c\u65f6\u4e0b\u8fbe\u6307\u4ee4\u3002\u6587\u7ae0\u8003\u5bdf LLM \u5728 &#8220;\u591a\u6b65\u957f\u7a0b\u89c4\u5212&#8221; \u4e0e &#8220;\u793e\u4ea4\u624b\u6bb5&#8221;\uff08\u5982\u5982\u4f55\u4e89\u53d6\u76df\u53cb\u3001\u5982\u4f55\u5236\u5b9a\u4fe1\u606f\u4e0d\u5bf9\u79f0\u7b56\u7565\uff09\u65b9\u9762\u7684\u7efc\u5408\u8868\u73b0\u3002\u4e0d\u4ec5\u4ec5\u662f\u8ba9\u6a21\u578b &#8220;\u6c42\u89e3&#8221; \u95ee\u9898\uff0c\u66f4\u662f\u8ba9\u6a21\u578b\u5728\u6709\u5176\u4ed6\u73a9\u5bb6\u5e72\u6270\u3001\u6216\u9700\u8981\u548c\u5176\u4ed6\u73a9\u5bb6\u6c9f\u901a\u7684\u573a\u666f\u4e2d\uff0c\u5b9e\u65f6\u5730\u8fdb\u884c\u7b56\u7565\u8c03\u6574\u3002\u8fd9\u5c31\u8981\u6c42 LLM \u8981\u5728\u8bed\u8a00\u80fd\u529b\u4e4b\u5916\uff0c\u5177\u5907\u591a\u6b65\u63a8\u7406\u548c\u5fc3\u667a\u6a21\u578b\uff08Theory of Mind\uff09\uff0c\u5e76\u80fd\u517c\u987e\u56e2\u961f \/ \u5bf9\u624b\u7684\u52a8\u673a\u3002<\/section>\n<section><\/section>\n<section><strong>\u5b9e\u9a8c\u7ed3\u679c\u4e0e\u5206\u6790: AI \u96c6\u4f53\u906d\u9047 &#8220;\u6ed1\u94c1\u5362&#8221;<\/strong><\/section>\n<section><\/section>\n<section><strong>\u6a21\u578b\u914d\u7f6e\u6982\u8ff0<\/strong><\/section>\n<section><\/section>\n<section>\u8be5\u6587\u8bc4\u4f30\u4e86\u5f53\u524d\u6d41\u884c\u7684\u95ed\u6e90\u548c\u5f00\u6e90\u5927\u8bed\u8a00\u6a21\u578b\uff1a<\/section>\n<section><\/section>\n<ul>\n<li>\n<section>\u95ed\u6e90\u6a21\u578b\uff1a\u5305\u542b OpenAI\uff08\u5982 o1, o3-mini, o1-mini, GPT-4o\u3001GPT-4o mini\u3001GPT-4-turbo \u7b49\uff09\u548c Anthropic \u7684 Claude 3.5\uff08Sonnet\/Haiku\uff09\uff0c\u5171 10 \u4e2a\u5546\u4e1a\u6a21\u578b\u3002<\/section>\n<\/li>\n<li>\n<section>\u5f00\u6e90\u6a21\u578b\uff1a\u6db5\u76d6 DeepSeek-R1\u3001Llama3 \u7cfb\u5217\u3001Qwen2.5-72B \u548c Mistral-7B \u7b49 7 \u4e2a\u6a21\u578b\u3002<\/section>\n<\/li>\n<\/ul>\n<section><a href=\"https:\/\/17aitech.com\/wp-content\/uploads\/2025\/03\/frc-b15337d15bd7a3c81fe338162f5ab3d0.png\" data-fancybox=\"images\" data-fancybox=\"gallery\"><img decoding=\"async\" src=\"https:\/\/17aitech.com\/wp-content\/uploads\/2025\/03\/frc-b15337d15bd7a3c81fe338162f5ab3d0.png\"><\/a><\/section>\n<section><em><sup>\u8868 1\uff1aPDDL\u3001\u7ade\u6280\u68cb\u76d8\u6e38\u620f\u548c\u5408\u4f5c\u6e38\u620f\u7684\u7ed3\u679c\u3002Plan Acc \u8868\u793a\u89c4\u5212\u51c6\u786e\u5ea6\u3002N-Step &#8221; \u8868\u793a N \u6b65\u524d\u77bb\u3002TTT\u3001C4\u3001CH \u662f\u4e09\u79cd\u7ade\u6280\u6e38\u620f\u3002WR \u4e0b\u6807\u8868\u793a\u4e13\u4e1a\u535a\u5f08\u5f15\u64ce\u5bf9\u6bcf\u4e2a LLM \u7684\u80dc\u7387\uff08%\uff09\u3002T3 \u4e0b\u6807\u663e\u793a LLM \u5728\u6240\u6709\u5bf9\u5c40\u4e2d\u7684\u68cb\u6b65\u5c5e\u4e8e top 3 choice \u7684\u767e\u5206\u6bd4\uff08%\uff09\u3002Hanabi \u5217\u663e\u793a 2-5 \u540d\u68cb\u624b\u53c2\u4e0e\u6e38\u620f\u7684\u5e73\u5747\u5f97\u5206\u3002<\/sup><\/em><\/section>\n<section><\/section>\n<section><strong>PDDL \u5b9e\u9a8c\u7ed3\u679c\u5206\u6790<\/strong><\/section>\n<section><a href=\"https:\/\/17aitech.com\/wp-content\/uploads\/2025\/03\/frc-ee37d9a39b71c9b5d41beef442c1b1ef.png\" data-fancybox=\"images\" data-fancybox=\"gallery\"><img decoding=\"async\" src=\"https:\/\/17aitech.com\/wp-content\/uploads\/2025\/03\/frc-ee37d9a39b71c9b5d41beef442c1b1ef.png\"><\/a><\/section>\n<section><em><sup>\u56fe 1\uff1ao1 \u7684\u51c6\u786e\u5ea6\u4e0e\u884c\u52a8\u7a7a\u95f4\u4e4b\u95f4\u7684\u5173\u7cfb\u3002\u5de6\u56fe\u63cf\u7ed8\u7684\u662f\u51c6\u786e\u7387\u4e0e\u5408\u6cd5\u884c\u52a8\u5e73\u5747\u6570\u91cf\u7684\u51fd\u6570\u5173\u7cfb\uff0c\u53f3\u56fe\u8003\u5bdf\u7684\u662f\u51c6\u786e\u7387\u4e0e\u72b6\u6001 &#8211; \u884c\u52a8\u7a7a\u95f4 \u5927\u5c0f\u7684\u5173\u7cfb\u3002<\/sup><\/em><\/section>\n<section><\/section>\n<ul>\n<li>\n<section>o1 \u5728\u7ecf\u5178\u89c4\u5212\u4e0a\u8868\u73b0\u6700\u4f73\uff0c\u4f46\u5728\u5927\u578b\u884c\u52a8 \/ \u72b6\u6001\u7a7a\u95f4\u4e2d\u4ecd\u6709\u660e\u663e\u4e0b\u964d\u3002<\/section>\n<\/li>\n<li>\n<section><strong>\u6838\u5fc3\u53d1\u73b0\uff1a\u6a21\u578b\u51c6\u786e\u7387\u4e0e\u72b6\u6001 &#8211; \u884c\u52a8\u7a7a\u95f4\u89c4\u6a21\u7684\u5173\u8054\u6027\u663e\u8457\u9ad8\u4e8e\u4e0e\u5408\u6cd5\u884c\u52a8\u5e73\u5747\u6570\u91cf\u7684\u5173\u8054\u6027\u3002<\/strong><\/section>\n<\/li>\n<li>\n<section>\u8ba4\u77e5\u89e3\u8bfb\uff1a\u6a21\u578b\u5728\u89c4\u5212\u65f6\u9700\u627f\u62c5\u672a\u6765\u5927\u91cf\u6f5c\u5728\u5206\u652f\u7684\u8ba4\u77e5\u8d1f\u62c5\uff0c\u5373\u4f7f\u6bcf\u4e00\u6b65\u4ec5\u6709\u5c11\u91cf\u6709\u6548\u9009\u62e9\u3002\u8fd9\u8868\u660e LLMs \u66f4\u6613\u53d7\u5168\u5c40\u590d\u6742\u6027\u5f71\u54cd\uff0c\u800c\u975e\u5355\u6b65\u51b3\u7b56\u9650\u5236\u3002<\/section>\n<\/li>\n<\/ul>\n<section><\/section>\n<section><strong>\u7ade\u6280\u68cb\u76d8\u6e38\u620f\u8868\u73b0<\/strong><\/section>\n<section><strong><br \/><\/strong><\/section>\n<section>\u5728 solver \u4e0e LLM \u7684\u5bf9\u51b3\u4e2d\uff0csolver \u51e0\u4e4e\u603b\u662f\u83b7\u80dc\u6216\u548c\u68cb\uff08\u8868 1\uff09\uff1a<\/section>\n<section><\/section>\n<ul>\n<li>\n<section>\u4e95\u5b57\u68cb\uff08Tic-Tac-Toe\uff09\uff1a\u8f83\u5f3a\u7684\u6a21\u578b\uff08\u5982 o1\uff0cClaude 3.5 Sonnet\uff09\u5076\u5c14\u80fd\u4e0e\u5b8c\u7f8e Solver \u8fbe\u6210\u5e73\u5c40\uff0c\u4f46\u591a\u6570\u60c5\u51b5\u4ecd\u843d\u8d25\u3002<\/section>\n<\/li>\n<li>\n<section>\u56db\u5b50\u68cb\uff08Connect Four\uff09\u4e0e\u56fd\u9645\u8c61\u68cb\uff08Chess\uff09\uff1a\u6e38\u620f Solver\uff08\u5982 Stockfish \u5f15\u64ce\uff09\u5bf9\u6240\u6709\u6d4b\u8bd5\u7684 LLMs \u4fdd\u6301 100% \u80dc\u7387\u3002<\/section>\n<\/li>\n<li>\n<section>LLMs \u5728\u56db\u8fde\u73af\u4e2d\u5076\u5c14\u80fd\u9009\u62e9\u6700\u4f18\u68cb\u6b65\uff0c\u4f46\u5728\u56fd\u9645\u8c61\u68cb\u4e2d\u51c6\u786e\u7387\u9aa4\u964d\uff0c\u51f8\u663e\u5176\u6df1\u5c42\u6218\u672f\u63a8\u7406\u4e0e\u5206\u652f\u6269\u5c55\u80fd\u529b\u4e0d\u8db3\u3002<\/section>\n<\/li>\n<\/ul>\n<section><a href=\"https:\/\/17aitech.com\/wp-content\/uploads\/2025\/03\/frc-1c48571a8ed8ffd251b04f55890716c1.png\" data-fancybox=\"images\" data-fancybox=\"gallery\"><img decoding=\"async\" src=\"https:\/\/17aitech.com\/wp-content\/uploads\/2025\/03\/frc-1c48571a8ed8ffd251b04f55890716c1.png\"><\/a><\/section>\n<section><strong>Hanabi \u591a\u4eba\u5408\u4f5c\u6e38\u620f\u5206\u6790<\/strong><\/section>\n<section><\/section>\n<section>\u591a\u4eba\u534f\u4f5c\u6311\u6218\uff1a<\/section>\n<section><\/section>\n<ul>\n<li>\n<section>\u5f53\u73a9\u5bb6\u6570\u91cf\u4ece 2 \u4eba\u589e\u81f3 5 \u4eba\u65f6\uff0c\u9876\u5c16\u6a21\u578b\uff08\u5982 o1\uff09\u7684\u5e73\u5747\u5f97\u5206\u4ece 16.4 \u964d\u81f3 14.2\uff0c\u8868\u660e\u591a\u4ee3\u7406\u534f\u8c03\u5bf9 LLMs \u7684\u7b56\u7565\u4e00\u81f4\u6027\u9020\u6210\u538b\u529b\u3002<\/section>\n<\/li>\n<li>\n<section>\u90e8\u5206\u6a21\u578b\uff08\u5982 o1-mini\u3001DeepSeek-R1\uff09\u5b83\u4eec\u7684\u9ad8\u65b9\u5dee\u4ee5\u53ca\u76f8\u5bf9\u8f83\u4f4e\u7684\u5e73\u5747\u5f97\u5206\u8868\u660e\u7f3a\u4e4f\u8db3\u591f\u7684\u7b56\u7565\u89c4\u5212\u80fd\u529b\u4ee5\u6709\u6548\u5730\u8fdb\u884c Hanabi \u6e38\u620f\uff0c\u65e0\u6cd5\u9002\u5e94\u590d\u6742\u534f\u4f5c\u573a\u666f\u3002<\/section>\n<\/li>\n<\/ul>\n<section><\/section>\n<section>\u4e0e\u4eba\u7c7b\u5bf9\u6bd4\uff1a<\/section>\n<section><\/section>\n<ul>\n<li>\n<section>\u4f5c\u8005\u56e2\u961f\u722c\u53d6\u5e76\u5206\u6790\u4e86 54,977 \u573a\u4eba\u7c7b Hanabi \u6e38\u620f\u6570\u636e\uff0c\u53d1\u73b0\u4eba\u7c7b\u5f97\u5206\u96c6\u4e2d\u5728 15-25 \u5206\u533a\u95f4\u3002<\/section>\n<\/li>\n<li>\n<section>\u5f53 LLMs \u5fc5\u987b\u534f\u8c03\u591a\u4e2a\u4e0d\u5b8c\u6574\u7684\u4fe1\u606f\u6e20\u9053\u5e76\u8ddf\u8e2a\u961f\u53cb\u4e0d\u65ad\u53d8\u5316\u7684\u77e5\u8bc6\u72b6\u6001\u65f6\uff0c\u6240\u6709 LLMs \u5747\u672a\u8fbe\u5230\u4eba\u7c7b\u5f97\u5206\u7684\u7b2c\u4e00\u56db\u5206\u4f4d\u6570\uff0c\u66b4\u9732\u5176 &#8220;\u793e\u4f1a\u667a\u80fd&#8221; \u666e\u904d\u4e0d\u8db3\u3002<\/section>\n<\/li>\n<\/ul>\n<section><a href=\"https:\/\/17aitech.com\/wp-content\/uploads\/2025\/03\/frc-15aaa7fb91a6043fb7de01087c17659c.png\" data-fancybox=\"images\" data-fancybox=\"gallery\"><img decoding=\"async\" src=\"https:\/\/17aitech.com\/wp-content\/uploads\/2025\/03\/frc-15aaa7fb91a6043fb7de01087c17659c.png\"><\/a><\/section>\n<section><em><sup>\u8868 2\uff1a4 \u4e2a\u73a9\u5bb6\u7684 Diplomacy \u6e38\u620f\u5b9e\u9a8c\u7ed3\u679c\u7ed3\u679c\uff0c\uff08\u7ed3\u679c\u8868\u793a\uff1a\u65e0\u534f\u5546 \/ \u6709\u534f\u5546\uff09\u5c55\u793a LLM \u4e0d\u540c\u7c7b\u522b\u6307\u4ee4\u7684\u6210\u529f\u7387\uff0c\u4ee5\u53ca\u6e38\u620f\u7ed3\u675f\u65f6\u4f9b\u5e94\u4e2d\u5fc3\uff08SC\uff09\u548c\u53d7\u63a7\u533a\u57df\uff08CR\uff09\u7684\u6570\u91cf\u3002\u53f3\u4fa7\u4ece\u5de6\u5230\u53f3\u4e3a\u8c08\u5224\u6d88\u606f\u7684\u8bc4\u6d4b\u6307\u6807\u3002\u57fa\u672c\u6280\u80fd\u6d4b\u8bd5\uff08BS\uff09\u663e\u793a\u7279\u5b9a\u6a21\u578b\u662f\u5426\u901a\u8fc7\u4e86\u5916\u4ea4\u6e38\u620f\u7684\u57fa\u7840\u6280\u80fd\u6d4b\u8bd5\u3002<\/sup><\/em><\/section>\n<section><\/section>\n<section><strong>Diplomacy \u6218\u7565\u8c08\u5224\u6e38\u620f\u5206\u6790\uff08\u8868 2\uff09<\/strong><\/section>\n<section><\/section>\n<section>\u57fa\u7840\u6280\u80fd\u6d4b\u8bd5\uff1a<\/section>\n<section><\/section>\n<ul>\n<li>\n<section>\u5728\u65e0\u8c08\u5224\u7684\u5355\u73a9\u5bb6\u6700\u7b80\u5355\u57fa\u51c6\u6d4b\u8bd5\u4e2d\uff0c\u4ec5 o1-preview\u3001o1 \u548c GPT-4o \u80fd\u5728 20 \u56de\u5408\u5185\u5360\u9886 18 \u4e2a\u8865\u7ed9\u4e2d\u5fc3\uff0c\u5176\u4ed6\u6a21\u578b\u5747\u5931\u8d25\uff0c\u53cd\u6620 LLMs \u5728\u957f\u671f\u6218\u7565\u89c4\u5212\u4e0a\u7684\u5c40\u9650\u6027\u3002<\/section>\n<\/li>\n<\/ul>\n<section><\/section>\n<section>\u7a7a\u95f4\u6307\u4ee4\u80fd\u529b\uff1a<\/section>\n<section><\/section>\n<ul>\n<li>\n<section>o1 \u5728\u57fa\u672c\u547d\u4ee4\uff08\u5982\u79fb\u52a8\u548c\u653b\u51fb\uff09\u7684\u6210\u529f\u7387\u4e0a\u8d85\u8fc7\u5176\u4ed6\u6a21\u578b\uff0820-30%\uff09\u3002<\/section>\n<\/li>\n<li>\n<section>\u5bf9\u4e8e\u9700\u8981\u591a\u6b65\u6216\u591a\u667a\u80fd\u4f53\u903b\u8f91\u7684\u590d\u6742\u884c\u52a8\uff08\u5982\u81ea\u6211\u652f\u63f4\u548c\u652f\u63f4\u4ed6\u4eba\uff09\uff0c\u6240\u6709\u6a21\u578b\u7684\u6027\u80fd\u90fd\u660e\u663e\u4e0b\u964d\u3002<\/section>\n<\/li>\n<\/ul>\n<section><\/section>\n<section>\u591a\u73a9\u5bb6\u573a\u666f\u8868\u73b0\uff1a<\/section>\n<section><\/section>\n<ul>\n<li>\n<section>\u968f\u7740\u53c2\u4e0e\u56fd\u6570\u91cf\u589e\u52a0\uff082-7\uff09\uff0cLLMs \u7684\u6307\u4ee4\u51c6\u786e\u6027\u3001\u653b\u51fb\u6210\u529f\u7387\u4e0e\u4f9b\u5e94\u4e2d\u5fc3\u6536\u76ca\u663e\u8457\u4e0b\u964d\uff0c\u8868\u660e\u5176\u96be\u4ee5\u5e94\u5bf9\u90e8\u5206\u53ef\u89c2\u6d4b\u6027\u4e0e\u52a8\u6001\u8054\u76df\u5173\u7cfb\u7684\u53e0\u52a0\u590d\u6742\u5ea6\u3002<\/section>\n<\/li>\n<li>\n<section>\u5bf9\u6a21\u578b\u8c08\u5224\u6d88\u606f\u7684\u5206\u6790\u8868\u660e\uff0c\u5927\u578b\u8bed\u8a00\u6a21\u578b\u5728\u8c08\u5224\u4e2d\u5c55\u73b0\u51fa\u5dee\u5f02\u5316\u7b56\u7565\uff1a\u6240\u6709\u6a21\u578b\u5747\u8868\u73b0\u51fa\u9ad8\u7b56\u7565\u4e00\u81f4\u6027\uff08\u6bd4\u7387 &gt; 0.90\uff09\uff0c\u5176\u4e2d o1 \u7684\u8bf4\u670d\u529b\u6700\u5f3a\uff08\u63a5\u53d7\u7387 65%\uff09\uff0c\u800c Claude 3.5 Haiku \u8f83\u5f31\uff0836%\uff09\u3002\u591a\u6570\u6a21\u578b\u503e\u5411\u63d0\u51fa\u4e92\u5229\u65b9\u6848\uff08\u5982 GPT-4-turbo\uff09\uff0c\u4f46 DeepSeek-R1 \u548c 3.5 Haiku \u66f4\u503e\u5411\u4e8e\u5f15\u53d1\u51b2\u7a81\u3002\u5c3d\u7ba1 GPT-4-turbo \u64c5\u957f\u6362\u4f4d\u601d\u8003\uff0cDeepSeek-R1 \u5584\u7528\u6761\u4ef6\u6027\u6218\u672f\uff0c\u4f46\u4ece\u7ed3\u679c\u6765\u770b\uff0c\u6240\u6709\u6a21\u578b\u5747\u7f3a\u4e4f\u4eba\u7c7b\u8c08\u5224\u8005\u7684\u7b56\u7565\u7075\u6d3b\u6027 \u2014\u2014 \u5c24\u5176\u5728\u590d\u6742\u793e\u4ea4\u4e92\u52a8\u4e2d\u8868\u73b0\u5c40\u9650\uff0c\u53cd\u6620\u51fa\u5f53\u524d AI \u7684\u793e\u4f1a\u667a\u80fd\u4ecd\u5904\u4e8e\u521d\u7ea7\u9636\u6bb5\u3002<\/section>\n<\/li>\n<\/ul>\n<section><\/section>\n<section>\u8c08\u5224\u7684\u8d1f\u9762\u5f71\u54cd\uff1a<\/section>\n<section><\/section>\n<ul>\n<li>\n<section>\u5f15\u5165\u8c08\u5224\u5f80\u5f80\u4f1a\u5bf9 o1 \u8fd9\u6837\u7684\u5f3a\u63a8\u7406\u6a21\u578b\u4ea7\u751f\u53cd\u76f4\u89c9\u7684\u5f71\u54cd\uff1a\u6267\u884c\u7684\u6307\u4ee4\u6570\u91cf\u4e0e\u6700\u7ec8\u5f97\u5206\uff08\u8865\u7ed9\u4e2d\u5fc3 \/ \u63a7\u5236\u533a\u57df\uff09\u5927\u5e45\u4e0b\u964d\uff0c\u800c\u90e8\u5206\u8f83\u5f31\u6a21\u578b\u53cd\u800c\u8868\u73b0\u7a33\u5b9a\u3002<\/section>\n<\/li>\n<li>\n<section>\u8fd9\u4e00\u7ed3\u679c\u8868\u660e\uff0c\u6fc0\u70c8\u7684\u793e\u4ea4\u4e92\u52a8\u53ef\u80fd\u4f1a\u7834\u574f\u5f3a\u63a8\u7406 LLM \u7684\u8ba1\u5212\u8fde\u8d2f\u6027\u548c\u63a8\u7406\u601d\u7ef4\u94fe\uff0c\u51f8\u663e\u5176 &#8220;\u601d\u7ef4\u94fe\u6269\u5c55&#8221; \u4e0e\u793e\u4f1a\u667a\u80fd\u95f4\u7684\u6f5c\u5728\u77db\u76fe\u3002<\/section>\n<\/li>\n<\/ul>\n<section><\/section>\n<section><strong>\u5b9e\u9a8c\u7ed3\u8bba\uff1aLLM \u7684\u75db\u70b9\u4e0e\u6311\u6218<\/strong><\/section>\n<section><\/section>\n<section>\u901a\u8fc7\u8fd9\u5957\u6db5\u76d6\u4ece\u57fa\u7840\u89c4\u5212\u5230\u591a\u667a\u80fd\u4f53\u7b56\u7565\u535a\u5f08\u7684\u8bc4\u6d4b\uff0c\u7814\u7a76\u8005\u5f97\u51fa\u4e86\u4e00\u4e9b\u5173\u952e\u7ed3\u8bba\uff1a<\/section>\n<section><\/section>\n<section><strong>\u7b80\u5355\u89c4\u5212\u8fd8\u884c\uff0c\u590d\u6742\u89c4\u5212\u65f6\u5927\u591a\u4e0d\u884c<\/strong><\/section>\n<section><\/section>\n<section>\u5f53\u72b6\u6001\u7a7a\u95f4\u5c0f\u3001\u53ef\u9009\u52a8\u4f5c\u5c11\u65f6\uff0cLLM \u53ef\u4ee5\u5b8c\u6210\u76f8\u5f53\u4e0d\u9519\u7684\u5355\u6b65\u6216\u77ed\u7a0b\u89c4\u5212\uff0c\u4f46\u4e00\u65e6\u95ee\u9898\u89c4\u6a21\u6269\u5f20\uff0c\u6216\u8005\u6e38\u620f\u8fdb\u5165\u4e2d\u540e\u671f\u51fa\u73b0\u5927\u91cf\u5206\u652f\uff0c\u6a21\u578b\u5c31\u5f88\u5feb\u51fa\u73b0\u591a\u6b65\u63a8\u7406\u74f6\u9888\uff0c\u751a\u81f3\u8f93\u51fa\u4e0d\u5408\u89c4\u5219\u7684\u884c\u52a8\u3002<\/section>\n<section><\/section>\n<section><strong>\u5927\u6a21\u578b\u80cc\u540e\u7684 &#8220;\u793e\u4ea4&#8221; \u4e0e &#8220;\u89c4\u5212&#8221; \u5f15\u64ce\u4ecd\u9700\u52a0\u5f3a<\/strong><\/section>\n<section><\/section>\n<section>\u672c\u6b21\u8bc4\u6d4b\u8868\u660e\u4e86\u5927\u6a21\u578b\u5728\u591a\u6b65\u51b3\u7b56\u4e0e\u4ed6\u4eba\u610f\u56fe\u5efa\u6a21\u65b9\u9762\u7684\u4e0d\u8db3\u3002\u672a\u6765\u82e5\u60f3\u771f\u6b63\u8ba9 LLM \u5728\u66f4\u590d\u6742\u3001\u66f4\u73b0\u5b9e\u7684\u591a\u667a\u80fd\u4f53\u534f\u540c\u573a\u666f\u53d1\u5149\u53d1\u70ed\uff0c\u6211\u4eec\u9700\u8981\u66f4\u5148\u8fdb\u7684\u5f3a\u5316\u5b66\u4e60\u6216\u591a\u667a\u80fd\u4f53\u8bad\u7ec3\u6846\u67b6\uff0c\u7ed3\u5408\u77e5\u8bc6\u56fe\u8c31\u3001\u8bb0\u5fc6\u6a21\u5757\u6216\u4e16\u754c\u6a21\u578b\u6765\u907f\u514d\u63a8\u7406\u94fe\u88ab\u8f7b\u6613\u6253\u65ad\u3002<\/section>\n<section><\/section>\n<section><strong>\u4e0d\u5b8c\u5168\u4fe1\u606f\u548c\u591a\u8df3\u63a8\u65ad\u662f\u786c\u4f24<\/strong><\/section>\n<section><\/section>\n<section>\u5728\u5982 Hanabi \u8fd9\u7c7b\u5e26 &#8220;\u9690\u542b\u4fe1\u606f&#8221; \u7684\u5408\u4f5c\u6e38\u620f\u4e2d\uff0c\u6a21\u578b\u9700\u8981\u901a\u8fc7\u961f\u53cb\u63d0\u793a\u6765\u63a8\u7406\u81ea\u5df1\u6301\u6709\u7684\u724c\u3002\u5b9e\u9a8c\u663e\u793a\uff0c\u5927\u591a\u6570 LLM \u4f9d\u65e7\u529b\u4e0d\u4ece\u5fc3\uff0c\u4e5f\u7f3a\u4e4f\u5bf9\u4ed6\u4eba\u601d\u7ef4\u8fdb\u884c\u591a\u8df3\u63a8\u7406\u7684\u7a33\u5b9a\u80fd\u529b\u3002<\/section>\n<section><\/section>\n<section><strong>\u4e0e\u4eba\u7c7b\u9ad8\u6c34\u5e73\u534f\u4f5c\u4ecd\u6709\u660e\u663e\u5dee\u8ddd<\/strong><\/section>\n<section><\/section>\n<section>\u5373\u4fbf\u662f\u8868\u73b0\u6700\u597d\u7684\u5927\u6a21\u578b\uff0c\u5728\u9700\u8981\u6df1\u5ea6\u5408\u4f5c\uff08\u5982 Hanabi\uff09\u6216\u591a\u65b9\u8c08\u5224\uff08\u5982 Diplomacy\uff09\u65f6\uff0c\u4ecd\u8fdc\u8fdc\u8fbe\u4e0d\u5230\u4eba\u7c7b\u73a9\u5bb6\u7684\u5e73\u5747\u6210\u7ee9\u3002\u8fd9\u4e5f\u4ece\u4e00\u4e2a\u4fa7\u9762\u8bf4\u660e\uff1a\u771f\u5b9e\u591a\u667a\u80fd\u4f53\u56e2\u961f\u51b3\u7b56\u4e2d\uff0c\u5927\u6a21\u578b\u8fd8\u9700\u8981\u5927\u91cf\u7684\u7ed3\u6784\u5316\u89c4\u5212\u6a21\u5757\u4e0e\u66f4\u4e30\u5bcc\u7684\u4ea4\u4e92\u8bb0\u5fc6 \/ \u63a8\u7406\u673a\u5236\u3002<\/section>\n<section><\/section>\n<section>\u4f5c\u8005\u7684\u9879\u76ee\u4e3b\u9875\u63d0\u4f9b\u4e86\u4e0d\u540c LLM \u4e4b\u95f4\u7684\u5bf9\u6218\u4ee5\u53ca\u6e38\u620f\u8f68\u8ff9\u7ec6\u8282\u548c\u4efb\u52a1\u7684\u53ef\u89c6\u5316\uff1ahttps:\/\/spinbench.github.io<\/section>\n<p>\u6587\u7ae0\u6765\u6e90\u4e8e\u4e92\u8054\u7f51:<a href=\"https:\/\/www.jiqizhixin.com\/articles\/2025-03-25-3\" target=\"_blank\">\u68cb\u76d8\u53d8\u6218\u573a\uff0c\u5927\u6a21\u578b\u5374\u5446\u4e86\uff1f\u666e\u6797\u65af\u987f\u3001UT Austin\u65b0\u57fa\u51c6SPIN-Bench\u66ddAI\u667a\u5546\u74f6\u9888<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"<p>\u6587\u7ae0\u6765\u6e90\u4e8e\u4e92\u8054\u7f51:\u68cb\u76d8\u53d8\u6218\u573a\uff0c\u5927\u6a21\u578b\u5374\u5446 [&hellip;]<\/p>\n","protected":false},"author":3,"featured_media":0,"comment_status":"open","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"site-sidebar-layout":"default","site-content-layout":"","ast-site-content-layout":"","site-content-style":"default","site-sidebar-style":"default","ast-global-header-display":"","ast-banner-title-visibility":"","ast-main-header-display":"","ast-hfb-above-header-display":"","ast-hfb-below-header-display":"","ast-hfb-mobile-header-display":"","site-post-title":"","ast-breadcrumbs-content":"","ast-featured-img":"","footer-sml-layout":"","theme-transparent-header-meta":"","adv-header-id-meta":"","stick-header-meta":"","header-above-stick-meta":"","header-main-stick-meta":"","header-below-stick-meta":"","astra-migrate-meta-layouts":"default","ast-page-background-enabled":"default","ast-page-background-meta":{"desktop":{"background-color":"var(--ast-global-color-4)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"ast-content-background-meta":{"desktop":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"footnotes":""},"categories":[27],"tags":[70,68],"class_list":["post-39649","post","type-post","status-publish","format-standard","hentry","category-news","tag-agent","tag-68"],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v26.4 - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>\u68cb\u76d8\u53d8\u6218\u573a\uff0c\u5927\u6a21\u578b\u5374\u5446\u4e86\uff1f\u666e\u6797\u65af\u987f\u3001UT Austin\u65b0\u57fa\u51c6SPIN-Bench\u66ddAI\u667a\u5546\u74f6\u9888 - \u4e00\u8d77AI\u6280\u672f<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/17aitech.com\/?p=39649\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\/\/17aitech.com\/?p=39649\",\"url\":\"https:\/\/17aitech.com\/?p=39649\",\"name\":\"\u68cb\u76d8\u53d8\u6218\u573a\uff0c\u5927\u6a21\u578b\u5374\u5446\u4e86\uff1f\u666e\u6797\u65af\u987f\u3001UT Austin\u65b0\u57fa\u51c6SPIN-Bench\u66ddAI\u667a\u5546\u74f6\u9888 - \u4e00\u8d77AI\u6280\u672f\",\"isPartOf\":{\"@id\":\"https:\/\/17aitech.com\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\/\/17aitech.com\/?p=39649#primaryimage\"},\"image\":{\"@id\":\"https:\/\/17aitech.com\/?p=39649#primaryimage\"},\"thumbnailUrl\":\"https:\/\/17aitech.com\/wp-content\/uploads\/2025\/03\/frc-84fab4aaea33ecaf86bc8a6a050a95b3.png\",\"datePublished\":\"2025-03-25T21:02:22+00:00\",\"author\":{\"@id\":\"https:\/\/17aitech.com\/#\/schema\/person\/60225458499e817ae0af73e67e440b9d\"},\"breadcrumb\":{\"@id\":\"https:\/\/17aitech.com\/?p=39649#breadcrumb\"},\"inLanguage\":\"zh-Hans\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/17aitech.com\/?p=39649\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"zh-Hans\",\"@id\":\"https:\/\/17aitech.com\/?p=39649#primaryimage\",\"url\":\"https:\/\/17aitech.com\/wp-content\/uploads\/2025\/03\/frc-84fab4aaea33ecaf86bc8a6a050a95b3.png\",\"contentUrl\":\"https:\/\/17aitech.com\/wp-content\/uploads\/2025\/03\/frc-84fab4aaea33ecaf86bc8a6a050a95b3.png\"},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/17aitech.com\/?p=39649#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"\u9996\u9875\",\"item\":\"https:\/\/17aitech.com\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"\u68cb\u76d8\u53d8\u6218\u573a\uff0c\u5927\u6a21\u578b\u5374\u5446\u4e86\uff1f\u666e\u6797\u65af\u987f\u3001UT Austin\u65b0\u57fa\u51c6SPIN-Bench\u66ddAI\u667a\u5546\u74f6\u9888\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/17aitech.com\/#website\",\"url\":\"https:\/\/17aitech.com\/\",\"name\":\"\u4e00\u8d77AI\u6280\u672f\",\"description\":\"\u8ba9AI\u77e5\u8bc6\u89e6\u624b\u53ef\u53ca\",\"alternateName\":\"\u4e00\u8d77AI\u6280\u672f\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/17aitech.com\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"zh-Hans\"},{\"@type\":\"Person\",\"@id\":\"https:\/\/17aitech.com\/#\/schema\/person\/60225458499e817ae0af73e67e440b9d\",\"name\":\"AI\u5c0f\u52a9\u624b\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"zh-Hans\",\"@id\":\"https:\/\/17aitech.com\/#\/schema\/person\/image\/\",\"url\":\"\/\/17aitech.com\/wp-content\/uploads\/2024\/04\/robot_3.png\",\"contentUrl\":\"\/\/17aitech.com\/wp-content\/uploads\/2024\/04\/robot_3.png\",\"caption\":\"AI\u5c0f\u52a9\u624b\"},\"description\":\"\u8fd9\u4e2a\u4eba\u5f88\u61d2\uff0c\u4ec0\u4e48\u90fd\u6ca1\u6709\u7559\u4e0b\uff5e\",\"url\":\"https:\/\/17aitech.com\/?page_id=33738&user=3\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"\u68cb\u76d8\u53d8\u6218\u573a\uff0c\u5927\u6a21\u578b\u5374\u5446\u4e86\uff1f\u666e\u6797\u65af\u987f\u3001UT Austin\u65b0\u57fa\u51c6SPIN-Bench\u66ddAI\u667a\u5546\u74f6\u9888 - \u4e00\u8d77AI\u6280\u672f","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/17aitech.com\/?p=39649","schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/17aitech.com\/?p=39649","url":"https:\/\/17aitech.com\/?p=39649","name":"\u68cb\u76d8\u53d8\u6218\u573a\uff0c\u5927\u6a21\u578b\u5374\u5446\u4e86\uff1f\u666e\u6797\u65af\u987f\u3001UT Austin\u65b0\u57fa\u51c6SPIN-Bench\u66ddAI\u667a\u5546\u74f6\u9888 - \u4e00\u8d77AI\u6280\u672f","isPartOf":{"@id":"https:\/\/17aitech.com\/#website"},"primaryImageOfPage":{"@id":"https:\/\/17aitech.com\/?p=39649#primaryimage"},"image":{"@id":"https:\/\/17aitech.com\/?p=39649#primaryimage"},"thumbnailUrl":"https:\/\/17aitech.com\/wp-content\/uploads\/2025\/03\/frc-84fab4aaea33ecaf86bc8a6a050a95b3.png","datePublished":"2025-03-25T21:02:22+00:00","author":{"@id":"https:\/\/17aitech.com\/#\/schema\/person\/60225458499e817ae0af73e67e440b9d"},"breadcrumb":{"@id":"https:\/\/17aitech.com\/?p=39649#breadcrumb"},"inLanguage":"zh-Hans","potentialAction":[{"@type":"ReadAction","target":["https:\/\/17aitech.com\/?p=39649"]}]},{"@type":"ImageObject","inLanguage":"zh-Hans","@id":"https:\/\/17aitech.com\/?p=39649#primaryimage","url":"https:\/\/17aitech.com\/wp-content\/uploads\/2025\/03\/frc-84fab4aaea33ecaf86bc8a6a050a95b3.png","contentUrl":"https:\/\/17aitech.com\/wp-content\/uploads\/2025\/03\/frc-84fab4aaea33ecaf86bc8a6a050a95b3.png"},{"@type":"BreadcrumbList","@id":"https:\/\/17aitech.com\/?p=39649#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"\u9996\u9875","item":"https:\/\/17aitech.com\/"},{"@type":"ListItem","position":2,"name":"\u68cb\u76d8\u53d8\u6218\u573a\uff0c\u5927\u6a21\u578b\u5374\u5446\u4e86\uff1f\u666e\u6797\u65af\u987f\u3001UT Austin\u65b0\u57fa\u51c6SPIN-Bench\u66ddAI\u667a\u5546\u74f6\u9888"}]},{"@type":"WebSite","@id":"https:\/\/17aitech.com\/#website","url":"https:\/\/17aitech.com\/","name":"\u4e00\u8d77AI\u6280\u672f","description":"\u8ba9AI\u77e5\u8bc6\u89e6\u624b\u53ef\u53ca","alternateName":"\u4e00\u8d77AI\u6280\u672f","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/17aitech.com\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"zh-Hans"},{"@type":"Person","@id":"https:\/\/17aitech.com\/#\/schema\/person\/60225458499e817ae0af73e67e440b9d","name":"AI\u5c0f\u52a9\u624b","image":{"@type":"ImageObject","inLanguage":"zh-Hans","@id":"https:\/\/17aitech.com\/#\/schema\/person\/image\/","url":"\/\/17aitech.com\/wp-content\/uploads\/2024\/04\/robot_3.png","contentUrl":"\/\/17aitech.com\/wp-content\/uploads\/2024\/04\/robot_3.png","caption":"AI\u5c0f\u52a9\u624b"},"description":"\u8fd9\u4e2a\u4eba\u5f88\u61d2\uff0c\u4ec0\u4e48\u90fd\u6ca1\u6709\u7559\u4e0b\uff5e","url":"https:\/\/17aitech.com\/?page_id=33738&user=3"}]}},"_links":{"self":[{"href":"https:\/\/17aitech.com\/index.php?rest_route=\/wp\/v2\/posts\/39649","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/17aitech.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/17aitech.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/17aitech.com\/index.php?rest_route=\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/17aitech.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=39649"}],"version-history":[{"count":0,"href":"https:\/\/17aitech.com\/index.php?rest_route=\/wp\/v2\/posts\/39649\/revisions"}],"wp:attachment":[{"href":"https:\/\/17aitech.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=39649"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/17aitech.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=39649"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/17aitech.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=39649"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}