{"id":1141445,"date":"2025-06-06T01:45:13","date_gmt":"2025-06-06T08:45:13","guid":{"rendered":"https:\/\/www.microsoft.com\/en-us\/research\/?post_type=msr-blog-post&p=1141445"},"modified":"2025-07-04T01:23:12","modified_gmt":"2025-07-04T08:23:12","slug":"value-compass-benchmarks","status":"publish","type":"msr-blog-post","link":"https:\/\/www.microsoft.com\/en-us\/research\/articles\/value-compass-benchmarks\/","title":{"rendered":"Value Compass Benchmarks\u81ea\u8fdb\u5316\u8bc4\u6d4b\u6846\u67b6\uff0c\u6df1\u5ea6\u5256\u6790\u5927\u6a21\u578b\u201c\u4e09\u89c2\u201d"},"content":{"rendered":"\n

\u7f16\u8005\u6309\uff1a\u5728 MBTI \u6d4b\u8bd5\u98ce\u9761\u7684\u5f53\u4e0b\uff0c\u4eba\u4eec\u70ed\u8877\u4e8e\u63a2\u5bfb\u81ea\u5df1\u7a76\u7adf\u662f\u5145\u6ee1\u6d3b\u529b\u7684\u201c\u5feb\u4e50\u5c0f\u72d7\u201d\uff0c\u8fd8\u662f\u654f\u611f\u7ec6\u817b\u7684\u201c\u6d41\u6cea\u732b\u732b\u201d\u3002\u5f53\u5927\u6a21\u578b\u9010\u6e10\u6210\u4e3a\u4eba\u4eec\u751f\u6d3b\u4e2d\u4e0d\u53ef\u6216\u7f3a\u7684\u52a9\u624b\u65f6\uff0c\u4f60\u662f\u5426\u4f1a\u597d\u5947\uff1a\u8fd9\u4e9b\u5927\u6a21\u578b\u6709\u7740\u600e\u6837\u7684\u7279\u6027\uff0c\u6216\u8005\u8bf4\uff0c\u5b83\u4eec\u79c9\u6301\u4f55\u79cd\u4e0d\u540c\u7684\u4ef7\u503c\u89c2\uff1f<\/p>\n\n\n\n

\u5fae\u8f6f\u4e9a\u6d32\u7814\u7a76\u9662\u6700\u65b0\u53d1\u5e03\u7684 Value Compass Benchmarks\uff08\u4ef7\u503c\u89c2\u7f57\u76d8\u8bc4\u4f30\u4e2d\u5fc3\uff09\uff0c\u53ef\u4ee5\u5e2e\u52a9\u7528\u6237\u4ee5\u66f4\u79d1\u5b66\u3001\u7cfb\u7edf\u548c\u53ef\u9760\u7684\u65b9\u5f0f\uff0c\u5bf9\u5927\u6a21\u578b\u7684\u4ef7\u503c\u89c2\u5c55\u5f00\u8bc4\u4f30\u3002\u540c\u65f6\uff0c\u7528\u6237\u901a\u8fc7\u8bbf\u95ee Value Compass Benchmarks \u7f51\u7ad9\uff0c\u53ef\u4ee5\u76f4\u63a5\u67e5\u770b\u4e3b\u6d41\u5927\u6a21\u578b\u7684\u4ef7\u503c\u89c2\u503e\u5411\uff0c\u5e76\u9009\u62e9\u6700\u9002\u5408\u81ea\u5df1\u7684\u5927\u6a21\u578b\u3002<\/p>\n\n\n\n


\n\n\n\n

\u5f53\u5927\u6a21\u578b\u9010\u6e10\u878d\u5165\u65e5\u5e38\u751f\u6d3b\uff0c\u5728\u4eab\u53d7\u6280\u672f\u5e26\u6765\u4fbf\u5229\u7684\u540c\u65f6\uff0c\u4f60\u662f\u5426\u601d\u8003\u8fc7\u8fd9\u4e9b\u5927\u6a21\u578b\u7684\u201c\u4e09\u89c2\u201d\u662f\u5426\u4e0e\u4f60\u5951\u5408\uff1f<\/p>\n\n\n\n

\u6709\u7684\u6a21\u578b\u5177\u6709\u521b\u9020\u529b\uff0c\u80fd\u591f\u6fc0\u53d1\u521b\u4f5c\u7075\u611f\uff1b\u6709\u7684\u6a21\u578b\u64c5\u4e8e\u603b\u7ed3\uff0c\u53ef\u4ee5\u4ece\u6d77\u91cf\u4fe1\u606f\u4e2d\u5feb\u901f\u63d0\u70bc\u6838\u5fc3\u8981\u70b9\uff1b\u6709\u7684\u6a21\u578b\u903b\u8f91\u7f1c\u5bc6\uff0c\u9762\u5bf9\u590d\u6742\u95ee\u9898\u80fd\u7ed9\u51fa\u6e05\u6670\u7684\u5206\u6790\u8def\u5f84\u2026\u2026<\/p>\n\n\n\n

\u5728\u8fd9\u4e2a\u767e\u5bb6\u4e89\u9e23\u7684\u5927\u6a21\u578b\u65f6\u4ee3\uff0c\u54ea\u4e2a\u6a21\u578b\u7684\u4ef7\u503c\u89c2\u548c\u4f60\u5951\u5408\u5ea6\u6700\u9ad8\uff1f<\/p>\n\n\n\n

\u4e3a\u4e86\u5e2e\u52a9\u7528\u6237\u66f4\u597d\u5730\u4e86\u89e3\u4e0d\u540c\u5927\u6a21\u578b\u5728\u4ef7\u503c\u89c2\u5c42\u9762\u7684\u8868\u73b0\uff0c\u5fae\u8f6f\u4e9a\u6d32\u7814\u7a76\u9662\u57fa\u4e8e\u6700\u65b0\u7814\u53d1\u7684\u4ef7\u503c\u89c2\u8bc4\u4f30\u6846\u67b6\uff0c\u63a8\u51fa\u4e86 Value Compass Benchmarks\uff08\u4ef7\u503c\u89c2\u7f57\u76d8\u8bc4\u4f30\u4e2d\u5fc3\uff09\u7f51\u7ad9 (opens in new tab)<\/span><\/a>\u3002\u8be5\u8bc4\u4f30\u6846\u67b6\u4f9d\u636e\u793e\u4f1a\u79d1\u5b66\u7406\u8bba\u8bbe\u8ba1\uff0c\u80fd\u591f\u7d27\u8ddf\u5927\u6a21\u578b\u7684\u53d1\u5c55\u6b65\u4f10\uff0c\u52a8\u6001\u3001\u79d1\u5b66\u3001\u7cfb\u7edf\u4e14\u5168\u9762\u5730\u8bc4\u4f30\u5927\u6a21\u578b\u7684\u4ef7\u503c\u89c2\u3002\u501f\u52a9 Value Compass Benchmarks \u7f51\u7ad9\uff0c\u7528\u6237\u53ef\u4ee5\u5bf9\u76ee\u524d\u5e02\u9762\u4e0a\u4e3b\u6d41\u7684\u5927\u6a21\u578b\u8fdb\u884c\u6df1\u5165\u7684\u4ef7\u503c\u89c2\u5206\u6790\uff0c\u4ece\u800c\u5728\u4e0d\u540c\u573a\u666f\u4e2d\u9009\u7528\u6700\u5951\u5408\u81ea\u8eab\u9700\u6c42\u7684\u5927\u6a21\u578b\uff0c\u4ee5\u66f4\u9ad8\u6548\u5730\u5e94\u5bf9\u5404\u7c7b\u95ee\u9898\u3002<\/p>\n\n\n\n

Value Compass Benchmarks \u7f51\u7ad9\u5730\u5740\uff1ahttps:\/\/valuecompass.github.io\/#\/benchmarks\/benchmarks (opens in new tab)<\/span><\/a><\/em><\/p>\n\n\n\n

\"graphical
\u56fe1\uff1aValue Compass Benchmarks \u7f51\u9875\u754c\u9762\u53ca\u529f\u80fd\u5c55\u793a<\/figcaption><\/figure>\n\n\n\n

\u4e86\u89e3\u5927\u6a21\u578b\u201c\u9690\u85cf\u201d\u7684\u4ef7\u503c\u89c2\u5c5e\u6027\uff0c\u9762\u4e34\u4e09\u5927\u56f0\u5883<\/h2>\n\n\n\n

\u968f\u7740\u5927\u6a21\u578b\u5728\u5404\u884c\u5404\u4e1a\u7684\u5e7f\u6cdb\u5e94\u7528\uff0c\u5176\u8f93\u51fa\u5185\u5bb9\u6240\u627f\u8f7d\u7684\u4ef7\u503c\u5bfc\u5411\u76f4\u63a5\u5f71\u54cd\u7740\u7528\u6237\u7684\u8ba4\u77e5\u3001\u5224\u65ad\u4e0e\u884c\u4e3a\u3002\u5982\u679c\u6a21\u578b\u7684\u4ef7\u503c\u89c2\u504f\u79bb\u4eba\u7c7b\u793e\u4f1a\u7684\u4e3b\u6d41\u4ef7\u503c\uff0c\u90a3\u4e48\u53ef\u80fd\u4f1a\u5e26\u6765\u8bef\u5bfc\u6027\u4fe1\u606f\u4f20\u64ad\u3001\u52a0\u5267\u793e\u4f1a\u504f\u89c1\u7b49\u98ce\u9669\u3002\u56e0\u6b64\uff0c\u7406\u89e3\u548c\u8bc4\u4f30\u5927\u6a21\u578b\u7684\u4ef7\u503c\u89c2\u503e\u5411\uff0c\u4e0d\u4ec5\u662f\u63d0\u5347\u5176\u5b89\u5168\u6027\u548c\u53ef\u63a7\u6027\u7684\u5173\u952e\u4e00\u6b65\uff0c\u4e5f\u662f\u9632\u6b62\u6ee5\u7528\u3001\u63a8\u52a8\u4eba\u5de5\u667a\u80fd\u5065\u5eb7\u53d1\u5c55\u7684\u5fc5\u8981\u524d\u63d0\u3002<\/p>\n\n\n\n

\u6b64\u5916\uff0c\u6bcf\u4e2a\u4e2a\u4f53\u6216\u4e0d\u540c\u7fa4\u4f53\u90fd\u6709\u81ea\u5df1\u72ec\u7279\u7684\u4ef7\u503c\u89c2\u548c\u9700\u6c42\uff0c\u5f53\u5927\u6a21\u578b\u6210\u4e3a\u4e0d\u53ef\u6216\u7f3a\u7684\u52a9\u624b\u65f6\uff0c\u4eba\u4eec\u4e5f\u66f4\u5e0c\u671b\u6a21\u578b\u4e0e\u81ea\u5df1\u7684\u6b65\u8c03\u4e00\u81f4\u3002\u4f8b\u5982\uff0c\u5bf9\u4e8e\u8ffd\u6c42\u81ea\u7531\u8868\u8fbe\u7684\u521b\u4f5c\u8005\uff0c\u9f13\u52b1\u5f00\u653e\u601d\u7ef4\u7684\u5927\u6a21\u578b\u66f4\u7b26\u5408\u9700\u6c42\uff1b\u5bf9\u4e8e\u5192\u9669\u7684\u521b\u65b0\u6d3e\uff0c\u5021\u5bfc\u6253\u7834\u5e38\u89c4\u3001\u652f\u6301\u591a\u5143\u78b0\u649e\u7684\u5927\u6a21\u578b\u4f1a\u662f\u66f4\u597d\u7684\u9009\u62e9\uff1b\u800c\u5bf9\u4e8e\u504f\u597d\u4f20\u7edf\u6587\u5316\u7684\u7fa4\u4f53\uff0c\u5f3a\u8c03\u5386\u53f2\u4f20\u627f\u7684\u5927\u6a21\u578b\u5219\u66f4\u80fd\u83b7\u5f97\u5171\u9e23\u3002\u901a\u8fc7\u4ef7\u503c\u89c2\u6d4b\u8bd5\uff0c\u7528\u6237\u53ef\u4ee5\u627e\u5230\u66f4\u201c\u5fd7\u540c\u9053\u5408\u201d\u7684\u6a21\u578b\uff0c\u66f4\u597d\u5730\u53d1\u6325\u5927\u6a21\u578b\u7684\u4f5c\u7528\uff0c\u5b9e\u73b0\u4e2a\u6027\u5316\u7684\u9700\u6c42\u3002<\/p>\n\n\n\n

\u7136\u800c\uff0c\u4ef7\u503c\u89c2\u4f5c\u4e3a\u4e00\u79cd\u5185\u5728\u800c\u590d\u6742\u7684\u5fc3\u7406\u7279\u5f81\uff0c\u5e76\u4e0d\u80fd\u50cf\u6027\u80fd\u6307\u6807\u90a3\u6837\u76f4\u89c2\u53ef\u89c1\u3001\u53ef\u4ee5\u88ab\u7b80\u5355\u5730\u91cf\u5316\u3002\u5fae\u8f6f\u4e9a\u6d32\u7814\u7a76\u9662\u7684\u7814\u7a76\u5458\u4eec\u6307\u51fa\uff0c\u5f53\u524d\u5bf9\u5927\u6a21\u578b\u4ef7\u503c\u89c2\u7684\u8bc4\u6d4b\u4e3b\u8981\u9762\u4e34\u4ee5\u4e0b\u4e09\u5927\u6838\u5fc3\u6311\u6218\uff1a<\/p>\n\n\n\n

\u5b9a\u4e49\u4e4b\u56f0\uff1a<\/strong>\u4eba\u7c7b\u4ef7\u503c\u89c2\u4f53\u7cfb\u5e9e\u6742\u4e14\u591a\u5143\uff0c\u6db5\u76d6\u81ea\u7531\u3001\u5e73\u7b49\u3001\u4fdd\u5b88\u3001\u5b89\u5168\u7b49\u591a\u4e2a\u7ef4\u5ea6\uff0c\u96be\u4ee5\u7528\u7b80\u5355\u7684\u6807\u7b7e\u6982\u62ec\u3002\u65e9\u671f\u7684\u8bc4\u4f30\u65b9\u6cd5\u591a\u805a\u7126\u4e8e\u201c\u662f\u5426\u4ea7\u751f\u6709\u5bb3\u5185\u5bb9\u201d\uff0c\u8fd9\u79cd\u5355\u4e00\u7ef4\u5ea6\u7684\u5224\u65ad\u65b9\u5f0f\uff0c\u5c31\u50cf\u4ec5\u51ed\u201c\u662f\u5426\u6492\u8c0e\u201d\u6765\u8bc4\u4ef7\u4e00\u4e2a\u4eba\u7684\u9053\u5fb7\u6c34\u5e73\u4e00\u6837\u7247\u9762\u4e14\u4e0d\u51c6\u786e\u3002\u5982\u4f55\u5728\u7eb7\u7e41\u590d\u6742\u7684\u4ef7\u503c\u4f53\u7cfb\u4e2d\u63d0\u70bc\u51fa\u53ef\u64cd\u4f5c\u3001\u53ef\u6d4b\u91cf\u7684\u8bc4\u4f30\u7ef4\u5ea6\uff0c\u51c6\u786e\u5b9a\u4e49\u8bc4\u4f30\u76ee\u6807\uff0c\u662f\u5f53\u524d\u4ef7\u503c\u89c2\u8bc4\u4f30\u9996\u5148\u8981\u89e3\u51b3\u7684\u6838\u5fc3\u95ee\u9898\u3002<\/p>\n\n\n\n

\u65f6\u6548\u4e4b\u56f0\uff1a<\/strong>\u73b0\u9636\u6bb5\u591a\u6570\u8bc4\u4f30\u4f9d\u8d56\u7279\u5b9a\u65f6\u95f4\u8282\u70b9\u7684\u9759\u6001\u6570\u636e\u96c6\u3002\u7136\u800c\uff0c\u968f\u7740\u5927\u6a21\u578b\u66f4\u65b0\u9891\u7387\u52a0\u5feb\uff0c\u51e0\u5929\u524d\u8fd8\u80fd\u6709\u6548\u533a\u5206\u6a21\u578b\u8868\u73b0\u7684\u6d4b\u8bd5\u9898\uff0c\u53ef\u80fd\u51e0\u5929\u540e\u5c31\u88ab\u6a21\u578b\u901a\u8fc7\u8bad\u7ec3\u201c\u719f\u80cc\u7b54\u6848\u201d\uff0c\u5931\u53bb\u8bc4\u4f30\u610f\u4e49\u3002\u6b64\u5916\uff0c\u516c\u5f00\u6570\u636e\u96c6\u672c\u8eab\u4e5f\u53ef\u80fd\u88ab\u7eb3\u5165\u8bad\u7ec3\u8bed\u6599\uff0c\u9020\u6210\u6570\u636e\u6c61\u67d3\uff0c\u8fdb\u4e00\u6b65\u524a\u5f31\u8bc4\u4f30\u7684\u771f\u5b9e\u6027\u548c\u6709\u6548\u6027\u3002\u90a3\u4e48\uff0c\u5982\u4f55\u6784\u5efa\u52a8\u6001\u3001\u53ef\u6301\u7eed\u66f4\u65b0\u7684\u8bc4\u4f30\u673a\u5236\uff1f<\/p>\n\n\n\n

\u77e5\u884c\u96be\u5408\u4e00\uff1a<\/strong>\u73b0\u6709\u8bc4\u4f30\u65b9\u6cd5\u5927\u591a\u91c7\u7528\u9009\u62e9\u9898\u6216\u95ee\u5377\u5f62\u5f0f\uff0c\u8ba9\u6a21\u578b\u201c\u81ea\u8ff0\u201d\u5176\u4ef7\u503c\u89c2\u503e\u5411\u3002\u8fd9\u79cd\u65b9\u5f0f\u867d\u7136\u4fbf\u4e8e\u5b9e\u65bd\uff0c\u4f46\u8bc4\u6d4b\u903b\u8f91\u4e0e\u771f\u5b9e\u5e94\u7528\u573a\u666f\u4e2d\u7684\u884c\u4e3a\u8868\u73b0\u5b58\u5728\u660e\u663e\u504f\u5dee\uff0c\u53ef\u80fd\u5bfc\u81f4\u6a21\u578b\u201c\u8bf4\u4e00\u5957\u505a\u4e00\u5957\u201d\u3002\u4f8b\u5982\uff0c\u4e00\u4e2a\u6a21\u578b\u53ef\u80fd\u5728\u201c\u516c\u5e73\u5c31\u4e1a\u201d\u95ee\u5377\u4e2d\u8868\u73b0\u79ef\u6781\uff0c\u5374\u5728\u751f\u6210\u62db\u8058\u6587\u6848\u65f6\u9690\u542b\u6027\u522b\u6216\u5e74\u9f84\u504f\u89c1\u3002\u5982\u4f55\u786e\u4fdd\u8bc4\u6d4b\u7ed3\u679c\u80fd\u771f\u5b9e\u53cd\u6620\u6a21\u578b\u5728\u5b9e\u9645\u4f7f\u7528\u4e2d\u7684\u884c\u4e3a\u8868\u73b0\uff1f\u5982\u4f55\u79d1\u5b66\u5ea6\u91cf\u548c\u89e3\u8bfb\u8bc4\u6d4b\u7ed3\u679c\uff1f\u4ecd\u662f\u5f53\u524d\u7814\u7a76\u4e2d\u7684\u8584\u5f31\u73af\u8282\u3002<\/p>\n\n\n\n

\u201c\u7528\u6237\u9700\u6c42\u4e0e\u4ef7\u503c\u89c2\u5177\u6709\u9c9c\u660e\u7684\u591a\u5143\u5316\u7279\u5f81\u3002\u4e0d\u540c\u6587\u5316\u80cc\u666f\u3001\u5e94\u7528\u573a\u666f\u751a\u81f3\u4e2a\u4f53\u7528\u6237\u4e4b\u95f4\uff0c\u5728\u4ef7\u503c\u89c2\u504f\u597d\u4e0a\u90fd\u53ef\u80fd\u5b58\u5728\u663e\u8457\u5dee\u5f02\u3002\u201d\u5fae\u8f6f\u4e9a\u6d32\u7814\u7a76\u9662\u7814\u7a76\u5458\u59da\u83c1\u8868\u793a\uff0c\u201c\u6211\u4eec\u5e0c\u671b\u901a\u8fc7\u4ef7\u503c\u89c2\u8bc4\u6d4b\uff0c\u5e2e\u52a9\u7528\u6237\u627e\u5230\u4e0e\u81ea\u5df1\u66f4\u5951\u5408\u7684\u5927\u6a21\u578b\uff0c\u4ece\u800c\u5b9e\u73b0\u66f4\u4e2a\u6027\u5316\u7684\u4f7f\u7528\u4f53\u9a8c\uff1b\u540c\u65f6\uff0c\u4e5f\u80fd\u5e2e\u52a9\u79d1\u7814\u4eba\u5458\u548c\u5f00\u53d1\u8005\u66f4\u6709\u9488\u5bf9\u6027\u5730\u4f18\u5316\u6a21\u578b\uff0c\u63a8\u52a8\u5927\u6a21\u578b\u4e0e\u4eba\u7c7b\u4ef7\u503c\u89c2\u66f4\u597d\u5730\u5bf9\u9f50\u3002\u201d<\/p>\n\n\n\n

Value Compass Benchmarks\u5b9e\u73b0\u66f4\u79d1\u5b66\u3001\u7cfb\u7edf\u3001\u5168\u9762\u7684\u5927\u6a21\u578b\u4ef7\u503c\u89c2\u8bc4\u4f30<\/h2>\n\n\n\n

\u4e3a\u5e94\u5bf9\u8fd9\u4e9b\u6311\u6218\uff0c\u5fae\u8f6f\u4e9a\u6d32\u7814\u7a76\u9662\u7684\u7814\u7a76\u5458\u4eec\u4ee5\u4eba\u7c7b\u57fa\u672c\u4ef7\u503c\u89c2\u4e3a\u8bc4\u4f30\u76ee\u6807\uff0c\u521b\u65b0\u6027\u5730\u5f00\u53d1\u4e86\u751f\u6210\u5f0f\u81ea\u8fdb\u5316\u8bc4\u4f30\u6846\u67b6\uff08Generative Self-Evolving Evaluation\uff09<\/strong>\uff0c\u5e76\u5b9a\u5236\u4e86\u591a\u7ef4\u5ea6\u52a0\u6743\u8bc4\u4f30\u6307\u6807\uff0c\u786e\u4fdd Value Compass Benchmarks \u80fd\u591f\u66f4\u79d1\u5b66\u3001\u7cfb\u7edf\u5730\u8bc4\u4f30\u5927\u6a21\u578b\u7684\u5185\u5728\u4ef7\u503c\u89c2\u3002<\/p>\n\n\n\n

\u9996\u5148\uff0c\u79fb\u690d\u6210\u719f\u7684\u793e\u4f1a\u79d1\u5b66\u7406\u8bba\u4f5c\u4e3a\u5ea6\u91cf\u8861<\/strong>\u3002\u7814\u7a76\u5458\u4eec\u501f\u9274\u4e86\u5fc3\u7406\u5b66\u548c\u793e\u4f1a\u79d1\u5b66\u4e2d\u5e7f\u6cdb\u8ba4\u53ef\u7684\u57fa\u672c\u4ef7\u503c\u89c2\u7406\u8bba\u6765\u5b9a\u4e49\u8bc4\u4f30\u76ee\u6807\u3002\u8fd9\u4e9b\u7406\u8bba\u6db5\u76d6\u4e00\u7ec4\u6570\u91cf\u6709\u9650\u4e14\u52a8\u673a\u72ec\u7acb\u7684\u4ef7\u503c\u7ef4\u5ea6\uff0c\u4e0d\u4ec5\u80fd\u591f\u7cfb\u7edf\u6027\u5730\u89e3\u91ca\u591a\u6837\u3001\u590d\u6742\u7684\u4eba\u7c7b\u884c\u4e3a\uff0c\u8fd8\u63ed\u793a\u4e86\u4e0d\u540c\u4ef7\u503c\u4f53\u7cfb\u95f4\u7684\u5173\u8054\u6027\u548c\u6f5c\u5728\u51b2\u7a81\u3002<\/p>\n\n\n\n