{"id":1102,"date":"2026-11-04T18:20:18","date_gmt":"2026-11-04T10:20:18","guid":{"rendered":"https:\/\/www.timedigital.cn\/new\/1102\/"},"modified":"2026-11-04T18:20:18","modified_gmt":"2026-11-04T10:20:18","slug":"evalship%e6%a3%80%e6%9f%a5%e5%a4%a7%e6%a8%a1%e5%9e%8b%e8%af%84%e6%b5%8b%e6%bc%8f%e6%b4%9e","status":"publish","type":"post","link":"https:\/\/www.timedigital.cn\/new\/1102\/","title":{"rendered":"Evalship\u68c0\u67e5\u5927\u6a21\u578b\u8bc4\u6d4b\u6f0f\u6d1e"},"content":{"rendered":"<div class=\"tdx-art\">\n<p class=\"tdx-lead\">Evalship \u5728 Hacker News \u5c55\u793a\u4e86\u4e00\u6b3e\u7528\u4e8e\u68c0\u67e5\u5927\u6a21\u578b\u8bc4\u6d4b\u7f3a\u9677\u7684\u5de5\u5177\u3002\u5176\u516c\u5f00\u793a\u4f8b\u53d1\u73b0\uff0c\u5ba2\u670d\u673a\u5668\u4eba\u7684\u56de\u590d\u5373\u4f7f\u8bfb\u51fa\u53e6\u4e00\u4f4d\u5ba2\u6237\u7684\u5730\u5740\uff0c\u4ecd\u80fd\u4ee5 <strong>0.80<\/strong> \u5206\u901a\u8fc7\u6d4b\u8bd5\uff1b\u95ee\u9898\u5b9a\u4f4d\u5728 <code>tests\/evals\/test_support_bot.py<\/code> <strong>\u7b2c 44 \u884c<\/strong>\u3002<\/p>\n<p>\u6d89\u4e8b\u8bc4\u6d4b\u91c7\u7528\u52a0\u6743\u603b\u5206\u8ba1\u7b97\u65b9\u5f0f\uff0c\u516c\u5f0f\u4e3a <code>score = 0.5 <em> helpfulness + 0.3 <\/em> tone + 0.2 * privacy<\/code>\u3002\u53cb\u597d\u5ea6\u3001\u8bed\u6c14\u548c\u9690\u79c1\u5206\u522b\u5360 0.5\u30010.3 \u548c 0.2 \u7684\u6743\u91cd\uff0c\u5176\u4e2d\u9690\u79c1\u9879\u5360\u4e94\u5206\u4e4b\u4e00\uff0c\u6ca1\u6709\u5355\u72ec\u89e6\u53d1\u6848\u4f8b\u5931\u8d25\u7684\u89c4\u5219\u3002\u603b\u5206\u8fbe\u5230\u901a\u8fc7\u7ebf\u540e\uff0c\u5373\u4f7f\u56de\u590d\u5305\u542b\u5ba2\u6237\u5730\u5740\uff0c\u6d4b\u8bd5\u4e5f\u4e0d\u4f1a\u81ea\u52a8\u5c06\u5176\u5224\u5b9a\u4e3a\u5931\u8d25\u3002<\/p>\n<p>\u516c\u5f00\u793a\u4f8b\u4e2d\u7684\u5ba2\u6237\u6570\u636e\u5c5e\u4e8e\u88ab\u6d4b\u8bd5\u5185\u5bb9\u3002\u9875\u9762\u4ee5\u8bfb\u51fa\u53e6\u4e00\u4f4d\u5ba2\u6237\u5730\u5740\u7684\u56de\u590d\u8bf4\u660e\u6f0f\u6d1e\uff0c\u6ca1\u6709\u62ab\u9732\u5177\u4f53\u5730\u5740\uff0c\u4e5f\u6ca1\u6709\u58f0\u79f0\u8be5\u5730\u5740\u5df2\u7ecf\u51fa\u73b0\u5728\u751f\u4ea7\u73af\u5883\u7684\u5ba2\u670d\u56de\u590d\u4e2d\u3002\u6848\u4f8b\u5c55\u793a\u7684\u662f\u81ea\u52a8\u8bc4\u6d4b\u5bf9\u9690\u79c1\u8fdd\u89c4\u7684\u6f0f\u68c0\uff0c\u5e76\u975e\u5df2\u7ecf\u53d1\u751f\u7684\u5ba2\u6237\u4fe1\u606f\u6cc4\u9732\u4e8b\u4ef6\u3002<\/p>\n<p>\u5bf9\u5f00\u53d1\u56e2\u961f\u800c\u8a00\uff0c\u7f3a\u9677\u53d1\u751f\u5728\u7528\u4e8e\u5224\u65ad\u5ba2\u670d\u673a\u5668\u4eba\u662f\u5426\u5408\u683c\u7684\u8bc4\u6d4b\u5c42\u3002\u6d4b\u8bd5\u4ee3\u7801\u628a\u56de\u590d\u8d28\u91cf\u4e0e\u9690\u79c1\u98ce\u9669\u5408\u5e76\u4e3a\u4e00\u4e2a\u5206\u6570\uff0c\u56e2\u961f\u4ec5\u4f9d\u636e\u603b\u5206\u51b3\u5b9a\u6848\u4f8b\u662f\u5426\u901a\u8fc7\uff0c\u5c31\u53ef\u80fd\u628a\u5305\u542b\u5ba2\u6237\u4fe1\u606f\u7684\u56de\u590d\u7eb3\u5165\u5408\u683c\u7ed3\u679c\u3002Evalship \u5c06\u8fd9\u7c7b\u95ee\u9898\u5f52\u4e3a\u4e25\u91cd\u8bc4\u6d4b\u6f0f\u6d1e\uff0c\u5e76\u628a\u5bf9\u5e94\u6587\u4ef6\u3001\u884c\u53f7\u548c\u4fee\u590d\u8981\u6c42\u4e00\u5e76\u5217\u51fa\u3002<\/p>\n<p>Evalship \u8fd8\u4e3a\u7f16\u7a0b\u667a\u80fd\u4f53\u751f\u6210\u9488\u5bf9\u8be5\u8bc4\u6d4b\u7684\u4fee\u590d\u6307\u4ee4\uff1a<\/p>\n<blockquote>\n<p>\u5355\u72ec\u8bc4\u4f30\u9690\u79c1\u9879\uff1b\u53ea\u8981\u56de\u590d\u53d1\u751f\u4efb\u4f55\u4fe1\u606f\u6cc4\u9732\uff0c\u65e0\u8bba\u5176\u4ed6\u9879\u76ee\u7684\u5f97\u5206\u5982\u4f55\uff0c\u90fd\u5e94\u5224\u5b9a\u8be5\u6848\u4f8b\u5931\u8d25\u3002<\/p>\n<\/blockquote>\n<p>\u8fd9\u9879\u6307\u4ee4\u8981\u6c42\u628a\u9690\u79c1\u68c0\u67e5\u4ece\u52a0\u6743\u603b\u5206\u4e2d\u62c6\u51fa\uff0c\u5e76\u4e3a\u4fe1\u606f\u6cc4\u9732\u8bbe\u7f6e\u72ec\u7acb\u7684\u5931\u8d25\u6761\u4ef6\u3002\u7f16\u7a0b\u667a\u80fd\u4f53\u53ef\u636e\u6b64\u4fee\u6539 <code>tests\/evals\/test_support_bot.py<\/code> \u4e2d\u7684\u8bc4\u6d4b\u903b\u8f91\uff0c\u4f7f\u51fa\u73b0\u5ba2\u6237\u5730\u5740\u7684\u56de\u590d\u65e0\u6cd5\u4f9d\u9760\u53cb\u597d\u5ea6\u548c\u8bed\u6c14\u5f97\u5206\u62b5\u6d88\u9690\u79c1\u6263\u5206\u3002<\/p>\n<p>\u8be5\u5de5\u5177\u7ed9\u51fa\u7684\u534f\u4f5c\u6d41\u7a0b\u4ee5\u62c9\u53d6\u8bf7\u6c42\u4f5c\u4e3a\u4fee\u590d\u8f7d\u4f53\u3002\u7f16\u7a0b\u667a\u80fd\u4f53\u63d0\u4ea4\u4ee3\u7801\u4fee\u6539\u540e\uff0cEvalship \u5bf9\u8bf7\u6c42\u4e2d\u7684\u4fee\u590d\u8fdb\u884c\u786e\u8ba4\uff0c\u8bc4\u6d4b\u6f0f\u6d1e\u7684\u68c0\u67e5\u3001\u4fee\u6539\u548c\u6838\u9a8c\u7531\u6b64\u5f62\u6210\u4e00\u5957\u9762\u5411\u4ee3\u7801\u4ed3\u5e93\u7684\u5904\u7406\u6d41\u7a0b\u3002<\/p>\n<p class=\"tdx-end\">\u00b7 END \u00b7<\/p>\n<\/div>\n","protected":false},"excerpt":{"rendered":"<p>Evalship \u5728 Hacker News \u5c55\u793a\u4e86\u4e00\u6b3e\u7528\u4e8e\u68c0\u67e5\u5927\u6a21\u578b\u8bc4\u6d4b\u7f3a\u9677\u7684\u5de5\u5177\u3002\u5176\u516c\u5f00\u793a\u4f8b\u53d1\u73b0\uff0c\u5ba2\u670d\u673a\u5668\u4eba\u7684\u56de\u590d\u5373\u4f7f\u8bfb\u51fa\u53e6\u4e00\u4f4d\u5ba2\u6237\u7684\u5730\u5740\uff0c\u4ecd\u80fd\u4ee5 0.80 \u5206\u901a\u8fc7\u6d4b\u8bd5\uff1b\u95ee\u9898\u5b9a\u4f4d\u5728 tests\/evals\/test_support_bot.py \u7b2c 44 \u884c\u3002<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[3],"tags":[],"class_list":["post-1102","post","type-post","status-publish","format-standard","hentry","category-3"],"_links":{"self":[{"href":"https:\/\/www.timedigital.cn\/new\/wp-json\/wp\/v2\/posts\/1102","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.timedigital.cn\/new\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.timedigital.cn\/new\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.timedigital.cn\/new\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.timedigital.cn\/new\/wp-json\/wp\/v2\/comments?post=1102"}],"version-history":[{"count":0,"href":"https:\/\/www.timedigital.cn\/new\/wp-json\/wp\/v2\/posts\/1102\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.timedigital.cn\/new\/wp-json\/wp\/v2\/media?parent=1102"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.timedigital.cn\/new\/wp-json\/wp\/v2\/categories?post=1102"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.timedigital.cn\/new\/wp-json\/wp\/v2\/tags?post=1102"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}