{"id":927,"date":"2025-08-27T09:22:00","date_gmt":"2025-08-27T01:22:00","guid":{"rendered":"http:\/\/www.applicationsofllm.com\/?p=927"},"modified":"2025-09-17T09:24:28","modified_gmt":"2025-09-17T01:24:28","slug":"%e5%a4%a7%e6%a8%a1%e5%9e%8b%e7%a2%b0%e5%88%b0%e7%9c%9f%e9%9a%be%e9%a2%98%e4%ba%86%ef%bc%8c%e6%b5%8b%e4%ba%86500%e9%81%93%ef%bc%8co3-pro%e4%bb%85%e9%80%9a%e8%bf%8715","status":"publish","type":"post","link":"http:\/\/www.applicationsofllm.com\/index.php\/2025\/08\/27\/%e5%a4%a7%e6%a8%a1%e5%9e%8b%e7%a2%b0%e5%88%b0%e7%9c%9f%e9%9a%be%e9%a2%98%e4%ba%86%ef%bc%8c%e6%b5%8b%e4%ba%86500%e9%81%93%ef%bc%8co3-pro%e4%bb%85%e9%80%9a%e8%bf%8715\/","title":{"rendered":"\u5927\u6a21\u578b\u78b0\u5230\u771f\u96be\u9898\u4e86\uff0c\u6d4b\u4e86500\u9053\uff0co3 Pro\u4ec5\u901a\u8fc715%"},"content":{"rendered":"\n<p>\u57fa\u51c6\u6d4b\u8bd5\u662f\u68c0\u9a8c\u5927\u6a21\u578b\u80fd\u529b\u7684\u4e00\u79cd\u65b9\u5f0f\uff0c\u4e00\u822c\u800c\u8a00\uff0c\u4e00\u4e2a\u6709\u7528\u7684\u57fa\u51c6\u65e2\u8981\u8db3\u591f\u96be\uff0c\u53c8\u8981\u8d34\u8fd1\u73b0\u5b9e\uff1a\u95ee\u9898\u65e2\u80fd\u6311\u6218\u524d\u6cbf\u6a21\u578b\uff0c\u53c8\u8981\u53cd\u6620\u771f\u5b9e\u4e16\u754c\u7684\u4f7f\u7528\u573a\u666f\u3002<\/p>\n\n\n\n<p>\u7136\u800c\uff0c\u73b0\u6709\u6d4b\u8bd5\u9762\u4e34\u7740\u300c\u96be\u5ea6\u2013\u771f\u5b9e\u6027\u300d\u7684\u77db\u76fe\uff1a\u4fa7\u91cd\u4e8e\u8003\u8bd5\u7684\u57fa\u51c6\u5f80\u5f80\u88ab\u4eba\u4e3a\u8bbe\u7f6e\u5f97\u5f88\u96be\uff0c\u4f46\u5b9e\u9645\u4ef7\u503c\u6709\u9650\uff1b\u800c\u57fa\u4e8e\u771f\u5b9e\u7528\u6237\u4ea4\u4e92\u7684\u57fa\u51c6\u53c8\u5f80\u5f80\u504f\u5411\u4e8e\u7b80\u5355\u7684\u9ad8\u9891\u95ee\u9898\u3002<\/p>\n\n\n\n<p>\u5728\u6b64\u80cc\u666f\u4e0b\uff0c\u6765\u81ea\u65af\u5766\u798f\u5927\u5b66\u3001\u534e\u76db\u987f\u5927\u5b66\u7b49\u673a\u6784\u7684\u7814\u7a76\u8005\u63a2\u7d22\u4e86\u4e00\u79cd\u622a\u7136\u4e0d\u540c\u7684\u65b9\u5f0f\uff1a\u5728\u672a\u89e3\u51b3\u7684\u95ee\u9898\u4e0a\u8bc4\u4f30\u6a21\u578b\u7684\u80fd\u529b\u3002<\/p>\n\n\n\n<p>\u4e0e\u4e00\u6b21\u6027\u6253\u5206\u7684\u9759\u6001\u57fa\u51c6\u4e0d\u540c\uff0c\u8be5\u7814\u7a76\u4e0d\u65ad\u6536\u96c6\u672a\u89e3\u51b3\u7684\u95ee\u9898\uff0c\u7136\u540e\u901a\u8fc7\u9a8c\u8bc1\u5668\u8f85\u52a9\u7b5b\u9009\u4e0e\u793e\u533a\u9a8c\u8bc1\u673a\u5236\uff0c\u5b9e\u73b0\u5bf9\u6a21\u578b\u7684\u6301\u7eed\u5f02\u6b65\u8bc4\u4f30\u3002<\/p>\n\n\n\n<p>\u5177\u4f53\u800c\u8a00\uff0c\u672c\u6587\u63d0\u51fa\u4e86 UQ\uff08Unsolved Questions\uff09\uff0c\u8fd9\u662f\u4e00\u4e2a\u7531 500 \u9053\u9898\u7ec4\u6210\u7684\u6d4b\u8bd5\u96c6\uff0c\u6db5\u76d6\u8ba1\u7b97\u673a\u7406\u8bba\u3001\u6570\u5b66\u3001\u79d1\u5e7b\u3001\u5386\u53f2\u7b49\u4e3b\u9898\uff0c\u7528\u4e8e\u8003\u5bdf\u6a21\u578b\u5728\u63a8\u7406\u3001\u4e8b\u5b9e\u51c6\u786e\u6027\u4ee5\u53ca\u6d4f\u89c8\u7b49\u65b9\u9762\u7684\u80fd\u529b\u3002UQ \u5728\u8bbe\u8ba1\u4e0a\u517c\u5177\u96be\u5ea6\u5927\u4e0e\u8d34\u8fd1\u771f\u5b9e\u4e24\u5927\u7279\u70b9\uff1a\u8fd9\u4e9b\u95ee\u9898\u5927\u591a\u662f\u4eba\u7c7b\u9047\u5230\u4f46\u5c1a\u672a\u89e3\u51b3\u7684\u96be\u9898\uff0c\u56e0\u6b64\u653b\u514b\u5b83\u4eec\u53ef\u76f4\u63a5\u4ea7\u751f\u73b0\u5b9e\u4ef7\u503c\u3002<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"422\" src=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-239-1024x422.jpg\" alt=\"\" class=\"wp-image-930\" srcset=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-239-1024x422.jpg 1024w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-239-300x124.jpg 300w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-239-768x317.jpg 768w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-239.jpg 1108w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<ul class=\"wp-block-list\">\n<li>\u8bba\u6587\u6807\u9898\uff1aUQ: Assessing Language Models on Unsolved Questions<\/li>\n\n\n\n<li>\u8bba\u6587\u5730\u5740\uff1ahttps:\/\/arxiv.org\/pdf\/2508.17580v1<\/li>\n\n\n\n<li>\u9879\u76ee\u5730\u5740\uff1ahttps:\/\/uq.stanford.edu\/<\/li>\n<\/ul>\n\n\n\n<p>\u603b\u7ed3\u800c\u8a00\uff0c\u672c\u6587\u8d21\u732e\u5982\u4e0b\uff1a<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>\u63d0\u51fa\u4e86 UQ \u6570\u636e\u96c6\u53ca\u5176\u6536\u96c6\u6d41\u7a0b\uff1a\u7ed3\u5408\u89c4\u5219\u8fc7\u6ee4\u5668\u3001\u5927\u8bed\u8a00\u6a21\u578b\u8bc4\u5ba1\u4ee5\u53ca\u4eba\u5de5\u5ba1\u6838\uff0c\u4ee5\u786e\u4fdd\u6700\u7ec8\u95ee\u9898\u7684\u8d28\u91cf\uff1b<\/li>\n\n\n\n<li>UQ-Validators\uff1a\u590d\u5408\u9a8c\u8bc1\u7b56\u7565\uff0c\u5229\u7528\u751f\u6210\u5668\u2013\u9a8c\u8bc1\u5668\u4e4b\u95f4\u7684\u80fd\u529b\u5dee\u8ddd\u6765\u6784\u5efa\u65e0\u771f\u503c\u9a8c\u8bc1\u7cfb\u7edf\uff08\u4e00\u822c\u800c\u8a00\u6a21\u578b\u9a8c\u8bc1\u80fd\u529b\u4f18\u4e8e\u751f\u6210\u80fd\u529b\uff09\uff0c\u5e76\u5bf9\u5019\u9009\u7b54\u6848\u8fdb\u884c\u9884\u7b5b\u9009\uff0c\u4ee5\u4fbf\u540e\u7eed\u4eba\u5de5\u5ba1\u6838\uff1b<\/li>\n\n\n\n<li>UQ-Platform\uff1a\u4e00\u4e2a\u5f00\u653e\u5e73\u53f0\uff0c\u8ba9\u4e13\u5bb6\u80fd\u591f\u5171\u540c\u9a8c\u8bc1\u95ee\u9898\u4e0e\u7b54\u6848\uff0c\u4ece\u800c\u5b9e\u73b0\u6301\u7eed\u7684\u3001\u5f02\u6b65\u7684\u3001\u793e\u533a\u9a71\u52a8\u7684\u8bc4\u4f30\u3002<\/li>\n<\/ul>\n\n\n\n<p>\u5b9e\u9a8c\u4e2d\uff0c\u8868\u73b0\u6700\u597d\u7684\u6a21\u578b\u4ec5\u5728 15% \u7684\u95ee\u9898\u4e0a\u901a\u8fc7\u4e86 UQ \u9a8c\u8bc1\uff0c\u800c\u521d\u6b65\u4eba\u5de5\u6838\u67e5\u5df2\u7ecf\u5728\u8fd9\u4e9b\u901a\u8fc7\u9a8c\u8bc1\u7684\u7b54\u6848\u4e2d\u8bc6\u522b\u51fa\u4e00\u4e9b\u6b63\u786e\u89e3\u7b54\u3002<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"333\" src=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-240-1024x333.jpg\" alt=\"\" class=\"wp-image-931\" srcset=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-240-1024x333.jpg 1024w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-240-300x97.jpg 300w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-240-768x250.jpg 768w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-240.jpg 1108w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<p><strong>\u6570\u636e\u96c6\u4ecb\u7ecd<\/strong><\/p>\n\n\n\n<p>UQ \u6570\u636e\u96c6\u7531 500 \u9053\u5177\u6709\u6311\u6218\u6027\u7684\u672a\u89e3\u51b3\u95ee\u9898\u7ec4\u6210\uff0c\u95ee\u9898\u6765\u6e90\u95ee\u7b54\u793e\u533a Stack Exchange\uff0c\u5e76\u4e14\u662f\u7ecf\u8fc7\u4e09\u8f6e\u7b5b\u9009\u5f97\u5230\u7684\u3002<\/p>\n\n\n\n<p>\u5728\u7b5b\u9009\u6d41\u7a0b\u4e0a\uff0c\u672c\u6587\u9996\u5148\u4eba\u5de5\u9009\u62e9\u4e86 80 \u4e2a Stack Exchange \u793e\u533a\uff08\u4f8b\u5982 Math Overflow\u3001Physics\uff09\uff0c\u5e76\u6293\u53d6\u5176\u4e2d\u672a\u89e3\u7b54\u7684\u95ee\u9898\uff0c\u5f97\u5230\u5927\u7ea6 300 \u4e07\u4e2a\u539f\u59cb\u5019\u9009\u95ee\u9898\u3002<\/p>\n\n\n\n<p>\u968f\u540e\uff0c\u8fdb\u5165\u591a\u9636\u6bb5\u7b5b\u9009\u6d41\u7a0b\u3002\u7b5b\u9009\u7684\u6bcf\u4e00\u9636\u6bb5\u90fd\u4f1a\u9010\u6b65\u7f29\u5c0f\u95ee\u9898\u6c60\uff1a\u57fa\u4e8e\u89c4\u5219\u7684\u7b5b\u9009\u5c06\u95ee\u9898\u7f29\u51cf\u81f3 33,916 \u4e2a\uff08\u5360\u539f\u59cb\u95ee\u9898\u6c60\u7684 1.13%\uff09\uff1b\u57fa\u4e8e\u5927\u8bed\u8a00\u6a21\u578b\u7684\u7b5b\u9009\u8fdb\u4e00\u6b65\u7f29\u51cf\u81f3 7,685 \u4e2a\uff08\u5360\u539f\u59cb\u7684 0.26%\uff09\uff1b\u6700\u7ec8\u901a\u8fc7\u4eba\u5de5\u5ba1\u6838\uff08\u5982\u5254\u9664\u6b8b\u7559\u7684\u91cd\u590d\u3001\u8fc7\u4e8e\u7b80\u5355\u3001\u504f\u9898\u6216\u8fdd\u53cd\u89c4\u5219\u7684\u95ee\u9898\uff09\uff0c\u5f97\u5230\u4e00\u4e2a\u7cbe\u5fc3\u6574\u7406\u7684 500 \u9053\u9898\u96c6\uff08\u5360\u539f\u59cb\u7684 0.02%\uff09\u3002<\/p>\n\n\n\n<p>\u968f\u7740\u95ee\u9898\u5728\u7b5b\u9009\u6d41\u7a0b\u4e2d\u9010\u6b65\u63a8\u8fdb\uff0c\u5b83\u4eec\u7684\u96be\u5ea6\u548c\u8d28\u91cf\u4e5f\u5728\u9010\u6e10\u63d0\u5347\u3002\u5c24\u5176\u662f\u57fa\u4e8e\u5927\u8bed\u8a00\u6a21\u578b\u7684\u7b5b\u9009\uff0c\u663e\u8457\u63d0\u9ad8\u4e86\u95ee\u9898\u7684\u96be\u5ea6\u3002<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"399\" src=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-238-1024x399.jpg\" alt=\"\" class=\"wp-image-929\" srcset=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-238-1024x399.jpg 1024w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-238-300x117.jpg 300w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-238-768x299.jpg 768w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-238.jpg 1108w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<p>\u6570\u636e\u96c6\u7ec4\u6210\u5982\u4e0b\u6240\u793a\uff0c\u4e3b\u8981\u5305\u542b\u79d1\u5b66\u7c7b\u95ee\u9898\uff0c\u5176\u6b21\u662f\u6280\u672f\u7c7b\u4e0e\u751f\u6d3b\u827a\u672f\u7c7b\u3002\u672c\u6587\u8fd8\u53d1\u73b0\u4e0d\u540c\u9886\u57df\u7684\u95ee\u9898\u80fd\u63a2\u6d4b\u6a21\u578b\u7684\u4e0d\u540c\u80fd\u529b\uff1a\u4f8b\u5982\u6570\u5b66\u95ee\u9898\u901a\u5e38\u9700\u8981\u5f00\u653e\u5f0f\u8bc1\u660e\uff0c\u800c\u79d1\u5e7b\u5947\u5e7b\u7c7b\u95ee\u9898\u5219\u504f\u91cd\u6d4f\u89c8\u68c0\u7d22\u80fd\u529b\uff08\u5982\u6839\u636e\u7247\u6bb5\u60c5\u8282\u8bc6\u522b\u4e66\u7c4d\u540d\u79f0\uff09\u3002<\/p>\n\n\n\n<p>\u4e00\u65e6\u67d0\u4e2a\u95ee\u9898\u88ab\u5224\u5b9a\u4e3a\u5df2\u89e3\u51b3\uff0c\u7814\u7a76\u8005\u5c31\u4f1a\u5728\u540e\u7eed\u7248\u672c\u4e2d\u5c06\u8be5\u95ee\u9898\u79fb\u9664\uff0c\u5e76\u7528\u65b0\u7684\u672a\u89e3\u51b3\u95ee\u9898\u66ff\u6362\u3002<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"440\" src=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-237-1024x440.jpg\" alt=\"\" class=\"wp-image-928\" srcset=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-237-1024x440.jpg 1024w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-237-300x129.jpg 300w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-237-768x330.jpg 768w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-237.jpg 1108w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<p><strong>UQ \u9a8c\u8bc1\u5668<\/strong><\/p>\n\n\n\n<p>\u867d\u7136 UQ \u6570\u636e\u96c6\u975e\u5e38\u5177\u6709\u4ef7\u503c\uff0c\u4f46\u8981\u5c06\u5176\u7528\u4f5c\u6a21\u578b\u6027\u80fd\u7684\u57fa\u51c6\uff0c\u4ecd\u9700\u914d\u5957\u7684\u8bc4\u5206\u6307\u6807\u3002\u7136\u800c\uff0c\u7531\u4e8e\u7f3a\u4e4f\u6807\u51c6\u7b54\u6848\uff0c\u65e0\u6cd5\u50cf\u8003\u8bd5\u57fa\u51c6\u90a3\u6837\u8fdb\u884c\u81ea\u52a8\u9a8c\u8bc1\u3002<\/p>\n\n\n\n<p>\u56e0\u6b64\uff0c\u672c\u6587\u8f6c\u5411\u65e0\u76d1\u7763\u9a8c\u8bc1\u5668\uff0c\u5373\u65e0\u9700\u6807\u51c6\u7b54\u6848\u3002\u7531\u4e8e\u672a\u89e3\u95ee\u9898\u5f80\u5f80\u6781\u5177\u6311\u6218\u6027\uff0c\u8fd9\u4e9b\u9a8c\u8bc1\u5668\u7684\u4e3b\u8981\u76ee\u6807\u5e76\u975e\u8bc1\u660e\u67d0\u4e2a\u5019\u9009\u7b54\u6848\u6b63\u786e\uff0c\u800c\u662f\u6392\u9664\u9519\u8bef\u7684\u5019\u9009\u7b54\u6848\uff1b\u56e0\u6b64\uff0c\u672c\u6587\u523b\u610f\u4f7f\u7528 validator\uff08\u9a8c\u8bc1\u5668\uff09\u4e00\u8bcd\uff0c\u800c\u975e judge \u6216 verifier\u3002<\/p>\n\n\n\n<p>\u9700\u8981\u7279\u522b\u6307\u51fa\u7684\u662f\uff0c\u7531\u4e8e\u7f3a\u5c11\u6807\u51c6\u7b54\u6848\uff0c\u8fd9\u7c7b\u9a8c\u8bc1\u5668\u672c\u8eab\u53ef\u80fd\u7ecf\u5e38\u51fa\u9519\uff0c\u4f46\u5b83\u4eec\u4ecd\u80fd\u5728\u540e\u7eed\u4eba\u5de5\u5ba1\u6838\u4e2d\u53d1\u6325\u8f85\u52a9\u4f5c\u7528\u3002<\/p>\n\n\n\n<p>\u636e\u4e86\u89e3\uff0c\u672c\u6587\u4e4b\u6240\u4ee5\u5f00\u53d1\u65e0\u9700\u6807\u51c6\u7b54\u6848\u7684\u9a8c\u8bc1\u5668\uff0c\u6838\u5fc3\u52a8\u673a\u5728\u4e8e\u8fd9\u6837\u4e00\u4e2a\u5047\u8bbe\uff1a\u5bf9\u96be\u9898\u5019\u9009\u7b54\u6848\u8fdb\u884c\u9a8c\u8bc1\u53ef\u80fd\u6bd4\u751f\u6210\u8fd9\u4e9b\u7b54\u6848\u66f4\u5bb9\u6613\u3002\u5b9e\u9a8c\u4e2d\u91c7\u7528\u4e86\u8fd9\u6837\u7684\u6d41\u7a0b\uff0c\u8ba9\u4e00\u7cfb\u5217\u80fd\u529b\u9012\u589e\u7684\u6a21\u578b\uff08\u4f8b\u5982 o3-mini \u2192 o4-mini \u2192 o3\uff09\u56de\u7b54\u8fd9 500 \u9053\u9898\uff0c\u8bb0\u5f55\u5b83\u4eec\u7684\u7b54\u9898\u51c6\u786e\u7387\uff1b\u63a5\u7740\uff0c\u8ba9\u6bcf\u4e2a\u6a21\u578b\u5728\u4e0d\u63a5\u89e6\u6807\u51c6\u7b54\u6848\u7684\u60c5\u51b5\u4e0b\uff0c\u9a8c\u8bc1\u5176\u4ed6\u6240\u6709\u6a21\u578b\u7ed9\u51fa\u7684\u7b54\u6848\uff1b\u6700\u540e\uff0c\u7528\u771f\u5b9e\u7b54\u6848\u5bf9\u8fd9\u4e9b\u9a8c\u8bc1\u7ed3\u8bba\u8fdb\u884c\u6253\u5206\uff0c\u8ba1\u7b97\u9a8c\u8bc1\u51c6\u786e\u7387\u3002<\/p>\n\n\n\n<p>\u56fe 5 \u5de6\u663e\u793a\uff1a\u968f\u7740\u6a21\u578b\u80fd\u529b\u7684\u63d0\u5347\uff0c\u5b83\u4eec\u5728\u9a8c\u8bc1\u51c6\u786e\u7387\u4e0a\u7684\u8fdb\u6b65\u901f\u5ea6\u660e\u663e\u5feb\u4e8e\u7b54\u9898\u51c6\u786e\u7387\u3002<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"506\" src=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-246-1024x506.jpg\" alt=\"\" class=\"wp-image-938\" srcset=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-246-1024x506.jpg 1024w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-246-300x148.jpg 300w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-246-768x380.jpg 768w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-246.jpg 1108w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<p>\u5b9e\u9a8c\u4e2d\u4f7f\u7528\u7684\u9a8c\u8bc1\u5668 pipeline\uff1a<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"194\" src=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-241-1024x194.jpg\" alt=\"\" class=\"wp-image-932\" srcset=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-241-1024x194.jpg 1024w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-241-300x57.jpg 300w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-241-768x146.jpg 768w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-241.jpg 1108w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<p><strong>\u5b9e\u9a8c\u53ca\u7ed3\u679c<\/strong><\/p>\n\n\n\n<p>\u5b9e\u9a8c\u8bc4\u4f30\u4e86 5 \u4e2a\u6a21\u578b\uff0c\u5305\u62ec o3\u3001o4-mini\u3001o3-mini\u3001Gemini 2.5 Pro \u548c Claude 3.7 Sonnet\u3002<\/p>\n\n\n\n<p>\u8868 1 \u7ed3\u679c\u663e\u793a\uff0c\u4e0e\u539f\u59cb\u57fa\u7ebf\u76f8\u6bd4\uff0c\u9a8c\u8bc1\u7b56\u7565\u80fd\u591f\u5b9e\u8d28\u6027\u5730\u63d0\u9ad8\u9a8c\u8bc1\u7684\u51c6\u786e\u7387\u548c\u7cbe\u5ea6\u3002\u4f8b\u5982\uff0c\u5bf9 Claude 3.7 Sonnet \u800c\u8a00\uff0c\u51c6\u786e\u7387\u4ece 21.6% \u63d0\u5347\u5230 73.2%\uff0c\u7cbe\u5ea6\u4ece 13.26% \u63d0\u5347\u5230 20%\uff0c\u4f46\u5f80\u5f80\u662f\u4ee5\u53ec\u56de\u7387\u4e0b\u964d\u4e3a\u4ee3\u4ef7\u3002<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"843\" src=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-242-1024x843.jpg\" alt=\"\" class=\"wp-image-933\" srcset=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-242-1024x843.jpg 1024w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-242-300x247.jpg 300w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-242-768x632.jpg 768w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-242.jpg 1108w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<p>\u4e3a\u4e86\u786e\u8ba4\u6700\u7ec8\u5f97\u5230\u7684\u6700\u4f73 UQ \u9a8c\u8bc1\u5668\u5bf9\u4eba\u7c7b\u8bc4\u5ba1\u8005\u6709\u5e2e\u52a9\uff0c\u8be5\u7814\u7a76\u9080\u8bf7\u82e5\u5e72\u8bc4\u5ba1\u5458\u5bf9 25 \u4e2a\u9a8c\u8bc1\u95ee\u9898\u8fdb\u884c\u8bc4\u5206\uff0c\u5224\u65ad\u5176\u7ed9\u51fa\u7684\u5224\u65ad\u7406\u7531\u94fe\u662f\u5426\u5728\u903b\u8f91\u4e0a\u6210\u7acb\u3002\u8868 2 \u663e\u793a\uff0c\u4eba\u7c7b\u8bc4\u5ba1\u4e0e\u9a8c\u8bc1\u5668\u7684\u4e00\u81f4\u7387\u53ca\u7406\u7531\u94fe\u7684\u51c6\u786e\u6027\u90fd\u5f88\u9ad8\uff0c\u8868\u660e\u8be5\u9a8c\u8bc1\u5668\u80fd\u4e3a\u4eba\u7c7b\u8bc4\u5ba1\u8005\u63d0\u4f9b\u6709\u6548\u652f\u6301\u3002<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"282\" src=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-244-1024x282.jpg\" alt=\"\" class=\"wp-image-935\" srcset=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-244-1024x282.jpg 1024w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-244-300x83.jpg 300w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-244-768x211.jpg 768w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-244.jpg 1108w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<p>\u5c06\u5927\u8bed\u8a00\u6a21\u578b\u7528\u4e8e\u7b54\u6848\u9a8c\u8bc1\u65f6\uff0c\u53e6\u4e00\u4e2a\u6311\u6218\u662f\u5b83\u4eec\u5e38\u5e38\u8868\u73b0\u51fa\u660e\u663e\u7684\u8bc4\u4f30\u504f\u89c1\u3002\u5f53\u7814\u7a76\u8005\u628a\u524d\u6cbf\u6a21\u578b\u76f4\u63a5\u5e94\u7528\u4e8e\u672c\u573a\u666f\u65f6\uff0c\u53d1\u73b0\u6240\u6709\u6a21\u578b\u5728\u8bc4\u4f30\u81ea\u8eab\u6216\u540c\u7cfb\u6a21\u578b\uff08\u5373\u540c\u4e00\u5f00\u53d1\u8005\u7684\u6a21\u578b\uff09\u65f6\uff0c\u90fd\u51fa\u73b0\u4e86\u8fc7\u5ea6\u4e50\u89c2\u73b0\u8c61\uff1a\u9884\u6d4b\u51fa\u7684\u6a21\u578b\u6027\u80fd\u8fdc\u9ad8\u4e8e\u5b9e\u9645\u6027\u80fd\uff0c\u5982\u56fe 7 \u6240\u793a\u3002<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Gemini \u660e\u663e\u504f\u5411\u81ea\u8eab\uff0c\u76f8\u5bf9\u4e8e\u5176\u4ed6\u6a21\u578b\u7ed9\u51fa\u663e\u8457\u66f4\u9ad8\u7684\u8bc4\u5206\uff1b<\/li>\n\n\n\n<li>Claude \u5bf9\u6240\u6709\u7b54\u6848\u6a21\u578b\uff08\u4e0d\u4ec5\u4ec5\u662f\u81ea\u8eab\uff09\u90fd\u8868\u73b0\u51fa\u8fc7\u5ea6\u4e50\u89c2\uff1b<\/li>\n\n\n\n<li>OpenAI \u7684 o \u7cfb\u5217\u6a21\u578b\u5219\u5bf9\u5176\u4ed6 o \u7cfb\u5217\u540c\u95e8\u6a21\u578b\u7ed9\u51fa\u8fc7\u9ad8\u8bc4\u4ef7\u3002<\/li>\n<\/ul>\n\n\n\n<p>\u968f\u7740\u6a21\u578b\u80fd\u529b\u9012\u589e\uff08o3-mini \u2192 o3\uff09\uff0c\u8fd9\u79cd\u504f\u89c1\u867d\u6709\u6240\u964d\u4f4e\uff0c\u4f46\u5e76\u672a\u5f7b\u5e95\u6d88\u9664\u3002<\/p>\n\n\n\n<p>\u672c\u6587\u8fdb\u4e00\u6b65\u53d1\u73b0\uff0c\u91c7\u7528\u590d\u5408\u9a8c\u8bc1\u5668\u80fd\u591f\u663e\u8457\u524a\u5f31\u7b54\u6848\u9a8c\u8bc1\u4e2d\u7684\u81ea\u6211\u504f\u89c1\u4e0e\u8fc7\u5ea6\u4e50\u89c2\u3002<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"967\" height=\"1024\" src=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-243-967x1024.jpg\" alt=\"\" class=\"wp-image-934\" srcset=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-243-967x1024.jpg 967w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-243-283x300.jpg 283w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-243-768x813.jpg 768w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-243.jpg 1108w\" sizes=\"auto, (max-width: 967px) 100vw, 967px\" \/><\/figure>\n\n\n\n<p>\u6700\u540e\uff0c\u672c\u6587\u8fd8\u53d1\u73b0\uff0c\u4e00\u4e2a\u66f4\u5f3a\u7684\u7b54\u6848\u751f\u6210\u6a21\u578b\u5e76\u4e0d\u4e00\u5b9a\u5c31\u662f\u66f4\u5f3a\u7684\u7b54\u6848\u9a8c\u8bc1\u6a21\u578b\u3002<\/p>\n\n\n\n<p>\u672c\u6587\u901a\u8fc7\u57fa\u7ebf\u63d0\u793a\u6cd5\u548c 3 \u8f6e\u8fed\u4ee3\u9a8c\u8bc1\u6d41\u7a0b\u7ed8\u5236\u4e86\u6a21\u578b\u5728 500 \u4e2a HLE \u95ee\u9898\u4e0a\u7684\u9a8c\u8bc1\u51c6\u786e\u7387\u4e0e\u7b54\u6848\u51c6\u786e\u7387\u5173\u7cfb\u56fe\u3002\u867d\u7136\u66f4\u597d\u7684\u7b54\u6848\u6027\u80fd\u901a\u5e38\u9884\u793a\u7740\u66f4\u597d\u7684\u9a8c\u8bc1\u6027\u80fd\uff08\u6574\u4f53\u5448\u4e0a\u5347\u8d8b\u52bf\uff09\uff0c\u4f46\u5e76\u975e\u7edd\u5bf9\u3002<\/p>\n\n\n\n<p>\u4f8b\u5982\uff1a\u5728\u6ca1\u6709\u6d41\u7a0b\u9a8c\u8bc1\u65f6\uff0co3 \u4f5c\u4e3a\u7b54\u6848\u6a21\u578b\u5f31\u4e8e Gemini 2.5 Pro\uff0c\u4f46\u4f5c\u4e3a\u9a8c\u8bc1\u6a21\u578b\u5374\u66f4\u5f3a\uff1b\u91c7\u7528\u6d41\u7a0b\u9a8c\u8bc1\u540e\uff0co3-mini \u4e0e Claude 3.7 Sonnet \u4e4b\u95f4\u89c2\u5bdf\u5230\u540c\u6837\u7684\u9006\u8f6c\u8d8b\u52bf\u3002\u6b64\u5916\uff0c\u5c3d\u7ba1 Claude 3.7 Sonnet \u5728\u7b54\u6848\u51c6\u786e\u7387\u4e0a\u663e\u8457\u843d\u540e\u4e8e Gemini 2.5 Pro\uff0c\u4f46\u5176\u57fa\u4e8e\u6d41\u7a0b\u9a8c\u8bc1\u7684\u8868\u73b0\u5374\u8d85\u8d8a\u4e86 Gemini 2.5 Pro \u7684\u57fa\u7ebf\u9a8c\u8bc1\u6027\u80fd\u3002<\/p>\n\n\n\n<figure class=\"wp-block-image size-full\"><img loading=\"lazy\" decoding=\"async\" width=\"581\" height=\"624\" src=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-147.png\" alt=\"\" class=\"wp-image-936\" srcset=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-147.png 581w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-147-279x300.png 279w\" sizes=\"auto, (max-width: 581px) 100vw, 581px\" \/><\/figure>\n","protected":false},"excerpt":{"rendered":"<p>\u57fa\u51c6\u6d4b\u8bd5\u662f\u68c0\u9a8c\u5927\u6a21\u578b\u80fd\u529b\u7684\u4e00\u79cd\u65b9\u5f0f\uff0c\u4e00\u822c\u800c\u8a00\uff0c\u4e00\u4e2a\u6709\u7528\u7684\u57fa\u51c6\u65e2\u8981\u8db3\u591f\u96be\uff0c\u53c8\u8981\u8d34\u8fd1\u73b0\u5b9e\uff1a\u95ee\u9898\u65e2\u80fd\u6311\u6218\u524d\u6cbf\u6a21\u578b\uff0c\u53c8 [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":937,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[10],"tags":[],"class_list":["post-927","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-10"],"_links":{"self":[{"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/posts\/927","targetHints":{"allow":["GET"]}}],"collection":[{"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/comments?post=927"}],"version-history":[{"count":1,"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/posts\/927\/revisions"}],"predecessor-version":[{"id":939,"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/posts\/927\/revisions\/939"}],"wp:featuredmedia":[{"embeddable":true,"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/media\/937"}],"wp:attachment":[{"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/media?parent=927"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/categories?post=927"},{"taxonomy":"post_tag","embeddable":true,"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/tags?post=927"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}