{"id":50990,"date":"2024-06-21T21:20:57","date_gmt":"2024-06-21T13:20:57","guid":{"rendered":"https:\/\/swarma.org\/?p=50990"},"modified":"2024-06-21T21:20:57","modified_gmt":"2024-06-21T13:20:57","slug":"%e6%8e%a2%e7%b4%a2%e5%a4%a7%e8%af%ad%e8%a8%80%e6%a8%a1%e5%9e%8b%e7%9a%84%e5%b9%b3%e7%ad%89%e3%80%81%e5%81%8f%e8%a7%81%e5%92%8c%e7%9c%9f%e5%ae%9e%e6%80%a7%e9%97%ae%e9%a2%98%e4%b8%a8%e5%91%a8%e5%85%ad","status":"publish","type":"post","link":"https:\/\/swarma.org\/?p=50990","title":{"rendered":"\u63a2\u7d22\u5927\u8bed\u8a00\u6a21\u578b\u7684\u5e73\u7b49\u3001\u504f\u89c1\u548c\u771f\u5b9e\u6027\u95ee\u9898\u4e28\u5468\u516d\u76f4\u64ad\u00b7AI+Social Science\u8bfb\u4e66\u4f1a"},"content":{"rendered":"
\n
\n
\n
\n
\n
\n
\n

<\/p>\n

\u56fe\u6e90\uff1aHugging Face\u3002https:\/\/huggingface.co\/blog\/evaluating-llm-bias<\/span><\/span><\/p>\n<\/section>\n

\n
\n
\n
\n
\n
\n

<\/section>\n<\/section>\n
\n
\n

\u5bfc\u8bed<\/strong><\/p>\n<\/section>\n<\/section>\n

\n

<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n
\n
\n
\u968f\u7740\u6587\u5b57\u548c\u56fe\u50cf\u751f\u6210\u5927\u6a21\u578b\u7684\u666e\u53ca\uff0c\u4eba\u5de5\u667a\u80fd\u6b63\u5728\u6084\u7136\u6539\u53d8\u6211\u4eec\u7684\u751f\u6d3b\u548c\u5de5\u4f5c\u65b9\u5f0f\u3002\u4f46\u662f\uff0c\u65b0\u7684\u95ee\u9898\u4e5f\u968f\u4e4b\u51fa\u73b0\uff1a\u8fd9\u4e9b\u751f\u6210\u5f0f\u5927\u6a21\u578b\u4e2d\u662f\u5426\u5b58\u5728\u793e\u4f1a\u6027\u98ce\u9669\uff0c\u4f8b\u5982\u504f\u89c1\u3001\u6b67\u89c6\uff1f<\/span><\/strong><\/section>\n

<\/span><\/strong><\/section>\n
\u672c\u6b21\u5206\u4eab\uff0c\u6211\u4eec\u9080\u8bf7\u5230\u52a0\u5dde\u5927\u5b66\u6d1b\u6749\u77f6\u5206\u6821\uff08UCLA\uff09\u8ba1\u7b97\u673a\u7cfb\u535a\u58eb\u751f\u3001\u4e9a\u9a6c\u900aAlexa\u751f\u6210\u5f0f\u4eba\u5de5\u667a\u80fd\uff08AGI\uff09\u7814\u7a76\u7ec4\u5b9e\u4e60\u751f\u2014\u2014\u4e07\u5955\u6b23\uff0c\u4e0e\u6211\u4eec\u8ba8\u8bba\u4e00\u4e0b\u95ee\u9898\uff1a\uff081\uff09\u751f\u6210\u5f0f\u5927\u6a21\u578b\u4e2d\u662f\u5426\u5b58\u5728\u504f\u89c1\u95ee\u9898\uff1f\uff082\uff09\u600e\u6837\u8861\u91cf\u6a21\u578b\u4e2d\u7684\u504f\u89c1\uff1f\uff083\uff09\u5982\u4f55\u89e3\u51b3\u6a21\u578b\u4e2d\u7684\u504f\u89c1\uff1f<\/span><\/strong><\/section>\n

<\/span><\/strong><\/section>\n
\u9664\u4e86\u5927\u8bed\u8a00\u6a21\u578b\u4e2d\u7684\u504f\u89c1\u95ee\u9898\uff0c\u672c\u6b21\u8fd8\u5206\u4eab\u4e86\u6a21\u578b\u504f\u89c1\u7684\u65b0\u5174\u7a76\u7814\u7a76\u9886\u57df\uff1a\u6587\u672c\u751f\u6210\u56fe\u7247\u5927\u6a21\u578b\u4e2d\u7684\u504f\u89c1\u95ee\u9898\u3002<\/span><\/strong><\/section>\n

<\/span><\/strong><\/section>\n
\u672c\u6b21\u5206\u4eab\u4e0d\u53ea\u662f\u5173\u4e8e\u6280\u672f\u95ee\u9898\u7684\u8ba8\u8bba\uff0c\u5b83\u8fd8\u9f13\u52b1\u6211\u4eec\u601d\u8003\u5982\u4f55\u5728\u5feb\u901f\u53d8\u5316\u7684AI\u9886\u57df\u4e2d\uff0c\u786e\u4fdd\u6280\u672f\u7684\u53d1\u5c55\u4e0e\u4eba\u7c7b\u7684\u4ef7\u503c\u89c2\u548c\u793e\u4f1a\u6b63\u4e49\u76f8\u534f\u8c03\u3002\u901a\u8fc7\u8fd9\u6b21\u6df1\u5165\u7684\u4ea4\u6d41\uff0c\u6211\u4eec\u5e0c\u671b\u6fc0\u53d1\u66f4\u591a\u5173\u4e8eAI\u6280\u672f\u793e\u4f1a\u8d23\u4efb\u7684\u601d\u8003\u4e0e\u8ba8\u8bba\u3002<\/span><\/strong><\/section>\n

<\/span><\/strong><\/section>\n
\u672c\u671f\u8bfb\u4e66\u4f1a\u65f6\u95f4\u4e3a\uff1a6\u670822\u65e510:30-12:30\uff08\u5468\u516d\uff09\uff0c\u76f4\u64ad\u62a5\u540d\u5165\u53e3\u89c1\u540e\u6587\u3002<\/span><\/strong><\/section>\n
<\/mp-common-videosnap><\/section>\n
<\/span><\/strong><\/section>\n<\/section>\n
\n
<\/strong><\/section>\n<\/section>\n
\n
<\/strong><\/section>\n<\/section>\n
\n
<\/strong><\/section>\n<\/section>\n
\n
<\/strong><\/section>\n<\/section>\n
\n
<\/strong><\/section>\n
<\/strong><\/section>\n
<\/strong><\/section>\n
<\/strong><\/section>\n
<\/strong><\/section>\n
<\/strong><\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n
\n
<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n

<\/span><\/section>\n
<\/span><\/section>\n

<\/span><\/section>\n

<\/span><\/section>\n

\n
\n
\n
\n
\n

<\/section>\n

<\/section>\n
\n
<\/svg><\/section>\n<\/section>\n<\/section>\n
\n

\u8be5\u5206\u4eab\u4e0e\u8bfb\u4e66\u4f1a\u6574\u4f53\u4e3b\u9898\u4e4b\u95f4\u7684\u5173\u7cfb<\/strong><\/span><\/strong><\/strong><\/p>\n<\/section>\n

\n

<\/section>\n

<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/h3>\n


<\/p>\n

\u5927\u8bed\u8a00\u6a21\u578b\u7684\u516c\u5e73\u3001\u504f\u89c1\u548c\u771f\u5b9e\u6027\u95ee\u9898\uff0c\u76f4\u63a5\u89e6\u53ca\u4e86\u6211\u4eec\u5982\u4f55\u5229\u7528\u8ba1\u7b97\u65b9\u6cd5\u6765\u89e3\u6790\u548c\u5e94\u5bf9AI\u6280\u672f\u4e2d\u56fa\u6709\u7684\u793e\u4f1a\u6027\u98ce\u9669\u3002\u8fd9\u4e0d\u4ec5\u662f\u6280\u672f\u6311\u6218\uff0c\u66f4\u662f\u793e\u4f1a\u79d1\u5b66\u7814\u7a76\u7684\u524d\u6cbf\u8bfe\u9898\u3002\u901a\u8fc7\u6df1\u5165\u5206\u6790\u5927\u578b\u8bed\u8a00\u6a21\u578b\u548c\u6587\u672c\u5230\u56fe\u50cf\u6a21\u578b\u4e2d\u7684\u504f\u89c1\u8868\u73b0\uff0c\u6211\u4eec\u6316\u6398\u4e86\u8ba1\u7b97\u793e\u4f1a\u79d1\u5b66\u5728\u4fc3\u8fdb\u6280\u672f\u4f26\u7406\u548c\u516c\u5e73\u6027\u65b9\u9762\u7684\u6f5c\u529b\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u5728\u8ba1\u7b97\u793e\u4f1a\u79d1\u5b66\u7684\u80cc\u666f\u4e0b\uff0c\u672c\u6b21\u5206\u4eab\u4e0d\u53ea\u662f\u5173\u4e8e\u6280\u672f\u95ee\u9898\u7684\u8ba8\u8bba\uff0c\u5b83\u8fd8\u9f13\u52b1\u6211\u4eec\u601d\u8003\u5982\u4f55\u5728\u5feb\u901f\u53d8\u5316\u7684AI\u9886\u57df\u4e2d\uff0c\u786e\u4fdd\u6280\u672f\u7684\u53d1\u5c55\u4e0e\u4eba\u7c7b\u7684\u4ef7\u503c\u89c2\u548c\u793e\u4f1a\u6b63\u4e49\u76f8\u534f\u8c03\u3002\u901a\u8fc7\u8fd9\u6b21\u6df1\u5165\u7684\u4ea4\u6d41\uff0c\u6211\u4eec\u5e0c\u671b\u6fc0\u53d1\u66f4\u591a\u5173\u4e8eAI\u6280\u672f\u793e\u4f1a\u8d23\u4efb\u7684\u601d\u8003\u4e0e\u8ba8\u8bba\u3002<\/span><\/section>\n

<\/section>\n


<\/strong><\/span><\/p>\n

\n
\n
\n
\n
\n

<\/section>\n

<\/section>\n
\n
<\/svg><\/section>\n<\/section>\n<\/section>\n
\n

\u7b80\u4ecb<\/strong><\/span><\/strong><\/strong><\/p>\n<\/section>\n

\n

<\/section>\n

<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/h3>\n


<\/p>\n

\u968f\u7740\u6587\u5b57\u548c\u56fe\u50cf\u751f\u6210\u5927\u6a21\u578b\u7684\u666e\u53ca\uff0c\u4eba\u5de5\u667a\u80fd\u6b63\u5728\u6084\u7136\u6539\u53d8\u6211\u4eec\u7684\u751f\u6d3b\u548c\u5de5\u4f5c\u65b9\u5f0f\u3002\u4f46\u662f\uff0c\u65b0\u7684\u95ee\u9898\u4e5f\u968f\u4e4b\u51fa\u73b0\uff1a\u8fd9\u4e9b\u751f\u6210\u5f0f\u5927\u6a21\u578b\u4e2d\u662f\u5426\u5b58\u5728\u793e\u4f1a\u6027\u98ce\u9669\uff0c\u4f8b\u5982\u504f\u89c1\u3001\u6b67\u89c6\uff1f\u6211\u4eec\u9700\u8981\u6ce8\u610f\u54ea\u4e9b\u751f\u6210\u5f0f\u5927\u6a21\u578b\u4e2d\u5b58\u5728\u7684\u793e\u4f1a\u79d1\u5b66\u95ee\u9898\uff1f\u662f\u5426\u6709\u65b9\u6cd5\u89e3\u51b3\uff1f\u672c\u6b21\u4e3b\u8bb2\u5185\u5bb9\u5206\u4e09\u90e8\u5206\u89e3\u7b54\u4e86\u8fd9\u4e09\u4e2a\u95ee\u9898\uff0c\u5e76\u4ecb\u7ecd\u4e86\u4e00\u7cfb\u5217\u76f8\u5173\u8bba\u6587\u3002\u9664\u4e86\u5927\u8bed\u8a00\u6a21\u578b\u4e2d\u7684\u504f\u89c1\u95ee\u9898\uff0c\u672c\u6b21\u8fd8\u5206\u4eab\u4e86\u6a21\u578b\u504f\u89c1\u7684\u65b0\u5174\u7a76\u7814\u7a76\u9886\u57df\uff1a\u6587\u672c\u751f\u6210\u56fe\u7247\u5927\u6a21\u578b\u4e2d\u7684\u504f\u89c1\u95ee\u9898\u3002\u5bf9\u4e8e\u8fd9\u4e2a\u95ee\u9898\uff0c\u6211\u4eec\uff081\uff09\u5206\u4eab\u4e86\u6700\u65b0\u7684\u8861\u91cf\u6587\u672c\u751f\u6210\u56fe\u7247\u5927\u6a21\u578b\u7684\u65b9\u6cd5\uff0c\uff082\uff09\u5206\u4eab\u4e86\u6700\u65b0\u7684\u5173\u4e8e\u6587\u672c\u751f\u6210\u56fe\u7247\u5927\u6a21\u578b\u4e2d\u504f\u89c1\u95ee\u9898\u7684\u7efc\u8ff0\u8bba\u6587\uff1b\uff083\uff09\u5206\u4eab\u4e86\u6307\u51fa\u73b0\u6709\u504f\u89c1\u89e3\u51b3\u65b9\u6848\u4e2d\u5b58\u5728\u7684\u771f\u5b9e\u6027\u95ee\u9898\uff08factuality\uff09\u7684\u6700\u65b0\u8bba\u6587\u3002\u901a\u8fc7\u5bf9\u8fd9\u4e9b\u6587\u7ae0\u7684\u89e3\u8bfb\uff0c\u672c\u6b21\u5206\u4eab\u65e8\u5728\u4e3a\u53c2\u4e0e\u8005\u63d0\u4f9b\u4e00\u4e9b\u5bf9\u4e8e\u8fd9\u4e2a\u65b0\u5174\u7814\u7a76\u65b9\u5411\u7684\u4fe1\u606f\u53ca\u8fdb\u5c55\uff0c\u5e76\u6307\u51fa\u4e86\u73b0\u6709\u7814\u7a76\u4e2d\u53ef\u4ee5\u6539\u8fdb\u7684\u90e8\u5206\uff0c\u4e3a\u611f\u5174\u8da3\u7684\u7814\u7a76\u8005\u6307\u51fa\u4e86\u5341\u5206\u6709\u6f5c\u529b\u7684\u7814\u7a76\u65b9\u5411\u3002<\/span><\/section>\n

<\/span><\/section>\n


<\/span><\/p>\n

\n
\n
\n
\n
\n

<\/section>\n

<\/section>\n
\n
<\/svg><\/section>\n<\/section>\n<\/section>\n
\n

\u53c2\u8003\u8bba\u6587<\/strong><\/span><\/strong><\/strong><\/p>\n<\/section>\n

\n

<\/section>\n

<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/h3>\n

 <\/p>\n

\u672c\u6b21\u5206\u4eab\u5206\u4e3a\u4e09\u90e8\u5206\uff1a\uff081\uff09\u751f\u6210\u5f0f\u5927\u6a21\u578b\u4e2d\u662f\u5426\u5b58\u5728\u504f\u89c1\u95ee\u9898\uff1f\uff082\uff09\u600e\u6837\u8861\u91cf\u6a21\u578b\u4e2d\u7684\u504f\u89c1\uff1f\uff083\uff09\u5982\u4f55\u89e3\u51b3\u6a21\u578b\u4e2d\u7684\u504f\u89c1\uff1f\u6bcf\u4e00\u90e8\u5206\u7684\u5206\u4eab\u90fd\u5305\u542b\u4e86\u76f8\u5173\u6700\u65b0\u8bba\u6587\u7684\u89e3\u8bfb\u3002\u4ee5\u4e0b\u662f\u8ba1\u5212\u5728\u6bcf\u4e00\u90e8\u5206\u5206\u4eab\u7684\u76f8\u5173\u8bba\u6587\u3002<\/span><\/p>\n

 \u2022 \u7b2c\u4e00\u90e8\u5206\uff1a\u751f\u6210\u5f0f\u5927\u6a21\u578b\u4e2d\u662f\u5426\u5b58\u5728\u504f\u89c1\u95ee\u9898\uff1f<\/span><\/p>\n