$\"\"$ <\/section>\n

\u672f\u8bed\u7ffb\u8bd1<\/section>\n<\/section>\n

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

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

\u5173\u4e8e\u6587\u7ae0<\/section>\n<\/section>\n<\/section>\n

\u539f\u6587\u6807\u9898<\/strong>\u00a0|\u00a0\u5927\u8bed\u8a00\u6a21\u578b\u65f6\u4ee3\u7684\u672f\u8bed\u7ffb\u8bd1<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/section>\n
\u539f\u6587\u520a\u767b<\/strong>\u00a0| \u300a\u4e2d\u56fd\u79d1\u6280\u672f\u8bed\u300b2024\u5e74\u7b2c3\u671f<\/span><\/section>\n
<\/section>\n
\u6587\u7ae0\u7248\u6743<\/strong>\u00a0| \u672c\u6587\u5df2\u83b7\u5f97\u539f\u4f5c\u8005\u6388\u6743\uff0c\u5982\u9700\u8f6c\u8f7d\uff0c\u53ef\u5173\u6ce8<\/span>\u672c\u53f7\uff0c\u5728\u540e\u53f0\u7559\u8a00<\/span><\/section>\n
\n
\n
02<\/span><\/section>\n
\u5185\u5bb9\u6458\u8981<\/section>\n<\/section>\n<\/section>\n
\n
\u5927\u8bed\u8a00\u6a21\u578b\u4e3a\u672f\u8bed\u5b66\u53d1\u5c55\u63d0\u4f9b\u4e86\u5343\u8f7d\u96be\u9022\u7684\u53d1\u5c55\u673a\u4f1a\uff0c\u6211\u56fd\u7684\u672f\u8bed\u5b66\u7814\u7a76\u5e94\u5f53\u7d27\u7d27\u6293\u4f4f\u8fd9\u4e2a\u673a\u4f1a\u3002\u5728\u5927\u8bed\u8a00\u6a21\u578b\u65f6\u4ee3\uff0c\u6211\u4eec\u5e94\u5f53\u628a\u672f\u8bed\u7684\u6c11\u65cf\u5316\u548c\u672f\u8bed\u7684\u56fd\u9645\u5316\u7ed3\u5408\u8d77\u6765\uff0c\u4ece\u800c\u63a8\u52a8\u6211\u56fd\u672f\u8bed\u5b66\u7684\u53d1\u5c55\u3002<\/span><\/span><\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n
\n
\n
\n
\n
\n
\n
\n
\n
\n
\n
01<\/strong><\/section>\n
02<\/strong><\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n
GPT\u7cfb\u5217\u4ea7\u54c1\u7684\u51fa\u73b0<\/strong><\/span><\/span><\/p>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n
\n
\n
\n
\n
2018\u5e74\uff0cOpenAI\u516c\u53f8\u5f00\u53d1\u4e86\u201c<\/span>\u57fa\u4e8e\u8f6c\u6362\u5668\u7684\u751f\u6210\u5f0f\u9884\u8bad\u7ec3\u6a21\u578b<\/span>\u201d\uff0c\u5176\u82f1\u6587\u672f\u8bed\u662fGenerative Pre-trained Transformer\uff0c\u7b80\u79f0GPT-1\u3002GPT-1\u7684\u63a8\u51fa\u63ed\u5f00\u4e86\u5927\u8bed\u8a00\u6a21\u578b(Large Language Model,LLM)\u7814\u5236\u7684\u5e8f\u5e55\u3002<\/span><\/span><\/p>\n
\n
GPT-1\u5229\u7528\u8f6c\u6362\u5668\u6a21\u578b(Transformer)\u7684\u7f16\u7801\u5668(encoder)\u548c\u89e3\u7801\u5668(decoder)\uff0c\u4ece\u8bed\u8a00\u5927\u6570\u636e\u4e2d\u83b7\u53d6\u4e86\u4e30\u5bcc\u7684\u8bcd\u6c47\u3001\u8bed\u6cd5\u548c\u8bed\u4e49\u77e5\u8bc6\uff0c\u5728\u8bed\u8a00\u751f\u6210\u4efb\u52a1\u4e0a\u8fbe\u5230\u4e86\u8f83\u9ad8\u6c34\u5e73\u3002<\/span><\/p>\n
\n
GPT-1\u4f7f\u7528Transformer\u8fdb\u884c\u8bad\u7ec3,\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d,\u4f7f\u7528\u6d77\u91cf\u7684\u81ea\u7136\u8bed\u8a00\u6587\u672c\u6570\u636e\u6765\u5b66\u4e60\u5355\u8bcd\u7684\u5d4c\u5165\u8868\u793a(word embedding expression)\u4ee5\u53ca\u4e0a\u4e0b\u6587\u4e4b\u95f4\u7684\u5173\u7cfb(context relation)\uff0c\u5f62\u6210\u77e5\u8bc6\u8868\u793a(knowledge representation)\u3002\u4e00\u65e6\u8bad\u7ec3\u5b8c\u6210,\u77e5\u8bc6\u8868\u793a\u5c31\u88ab\u7f16\u7801\u5728\u795e\u7ecf\u7f51\u7edc\u7684\u53c2\u6570\u4e2d\uff0c\u5229\u7528\u8fd9\u4e9b\u53c2\u6570\u53ef\u4ee5\u751f\u6210\u56de\u7b54\u3002\u5f53\u7528\u6237\u63d0\u51fa\u95ee\u9898\u65f6\uff0c\u795e\u7ecf\u7f51\u7edc\u5c31\u6839\u636e\u5df2\u7ecf\u5b66\u4e60\u5230\u7684\u77e5\u8bc6\uff0c\u628a\u7b54\u6848\u53cd\u9988\u7ed9\u7528\u6237\u3002<\/span><\/p>\n
\n
OpenAI\u516c\u53f8\u5206\u522b\u4e8e2018\u30012019\u5e74\u5f00\u53d1\u51fa\u4e86GPT-1\u548cGPT-2\uff0c\u4e8e2020\u5e745\u6708\u5f00\u53d1\u51fa\u4e86GPT-3\uff0c2020\u5e747\u6708\u7814\u5236\u4e86GPT-3\u4e2d\u7684davinci, curie, babbage\u4e09\u4e2a\u6a21\u578b,2022\u5e743\u6708\u7814\u5236\u4e86<\/span>InstructGPT<\/span>\uff0c\u8fdb\u884c\u6587\u672c\u548c\u4ee3\u7801\u7684\u8bed\u8a00\u6a21\u578b\u8bad\u7ec3\uff0c\u7814\u5236\u51faGPT-3.5(\u5982\u4e0b\u56fe1\u6240\u793a)\u540e\u63a5\u7740\u8fdb\u884c\u6709\u76d1\u7763\u5fae\u8c03(Supervised Fine Tuning, SFT)\u548c\u57fa\u4e8e\u4eba\u7c7b\u53cd\u9988\u7684\u5f3a\u5316\u5b66\u4e60(Reinforcement Learning from Human Feedback,RLHF)<\/span><\/span>\uff0c<\/span>\u4e4b\u540e\u4e8e2022\u5e7411\u6708\u63a8\u51fa<\/span>ChatGPT<\/span>\u3002<\/span><\/span><\/span><\/p>\n
\n
$\"\"$ <\/section>\n
ChatGPT\u7684\u63a8\u51fa\u5f15\u8d77\u4e86\u53f2\u65e0\u524d\u4f8b\u7684\u8f70\u52a8\u3002\u6210\u5343\u4e0a\u4e07\u7684\u7528\u6237\u4ece\u4e0d\u540c\u89d2\u5ea6\u5bf9\u5b83\u8fdb\u884c\u4e86\u5e94\u7528\u4f53\u9a8c\u3002ChatGPT\u57282022\u5e7411\u670830\u65e5\u63a8\u51fa\u4ec55\u5929\u65f6,\u6ce8\u518c\u7528\u6237\u5c31\u8d85\u8fc7\u767e\u4e07;\u63a8\u51fa\u4e24\u4e2a\u6708\u7684\u65f6\u5019,\u6708\u6d3b\u8dc3\u7528\u6237\u8d85\u8fc71\u4ebf\u3002\u77ed\u89c6\u9891\u793e\u4ea4\u5a92\u4f53\u5e73\u53f0TikTok\u6708\u6d3b\u8dc3\u7528\u6237\u8d85\u8fc71\u4ebf\u7528\u4e869\u4e2a\u6708\u65f6\u95f4,\u63a8\u7279(Twitter)\u6708\u6d3b\u8dc3\u7528\u6237\u8d85\u8fc71\u4ebf\u7528\u4e8690\u4e2a\u6708\u65f6\u95f4\uff0cChatGPT\u6253\u7834\u4e86\u5386\u53f2\u8bb0\u5f55\uff0c\u9065\u9065\u9886\u5148\uff0c\u5f15\u8d77\u4e86\u5168\u7403\u7f51\u6c11\u7684\u5e7f\u6cdb\u5173\u6ce8\uff0c<\/span>\u5728\u5927\u8bed\u8a00\u6a21\u578b\u65f6\u4ee3\u6380\u8d77\u4e86\u4e00\u573a\u6ce2\u6f9c\u58ee\u9614\u3001\u53f2\u65e0\u524d\u4f8b\u7684\u6d77\u5578^{[1<\/span><\/sup>^{]<\/span><\/sup>\u3002<\/span><\/span><\/p>\n}}
\n
\n
\n
\n
\n
\n
\n
\n
02<\/strong><\/section>\n
02<\/strong><\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n
ChatGPT\u7684\u4e3b\u8981\u5de5\u4f5c\u673a\u5236<\/span><\/strong><\/span><\/section>\n<\/section>\n<\/section>\n
<\/section>\n
ChatGPT\u662f\u4e00\u79cd\u5bf9\u8bdd\u573a\u666f\u4f18\u5316\u8bed\u8a00\u6a21\u578b(optimizing language models for dialogue)\u3002ChatGPT\u6bd4GPT-3\u66f4\u8fdb\u4e00\u6b65,\u8fdb\u5316\u5230\u5177\u5907\u6267\u884c\u81ea\u7136\u8bed\u8a00\u6307\u4ee4\u7684\u80fd\u529b\uff0c\u7528\u6237\u4e0d\u5fc5\u7ed9\u51fa\u793a\u4f8b,\u53ea\u8981\u4f7f\u7528\u81ea\u7136\u8bed\u8a00\u7ed9\u51fa\u6307\u4ee4\uff0cChatGPT\u5c31\u53ef\u4ee5\u7406\u89e3\u7528\u6237\u7684\u610f\u56fe\u3002\u4f8b\u5982\uff0c<\/span>\u7528\u6237\u53ea\u8981\u76f4\u63a5\u7528\u81ea\u7136\u8bed\u8a00\u544a\u8bc9ChatGPT\u628a\u67d0\u4e2a\u82f1\u8bed\u5355\u8bcd\u8bd1\u6210\u6cd5\u8bed\uff0c<\/span>ChatGPT\u5c31\u53ef\u4ee5\u6267\u884c\u5e76\u7ed9\u51fa\u7ffb\u8bd1\u7ed3\u679c\u3002ChatGPT\u53ef\u4ee5\u6839\u636e\u4e0a\u4e0b\u6587\u63d0\u793a\uff0c\u81ea\u52a8\u7406\u89e3\u5e76\u6267\u884c\u5404\u7c7b\u4efb\u52a1\uff0c\u4e0d\u5fc5\u66f4\u65b0\u6a21\u578b\u7684\u53c2\u6570\u6216\u67b6\u6784\u3002<\/span><\/p>\n
\n
ChatGPT\u5229\u7528Transformer\u6a21\u578b\uff0c<\/span>\u4ece\u8bed\u8a00\u5927\u6570\u636e\u4e2d\u83b7\u53d6\u4e30\u5bcc\u7684\u8bed\u8a00\u77e5\u8bc6\uff0c<\/span>\u5728\u8bed\u8a00\u751f\u6210\u4efb\u52a1\u4e0a\u8fbe\u5230\u4e86\u76f8\u5f53\u9ad8\u7684\u6c34\u5e73\uff0c<\/span>\u6210\u4e3a\u4e86\u5927\u8bed\u8a00\u6a21\u578b\u65f6\u4ee3\u6700\u91cd\u8981\u7684\u795e\u7ecf\u7f51\u7edc\u6a21\u578b\u3002<\/span><\/p>\n
\n
ChatGPT\u7684\u8bad\u7ec3\u8bed\u6599\u9ad8\u8fbe100\u4ebf\u4e2a\u53e5\u5b50\uff0c<\/span>\u5305\u542b\u7ea65000\u4ebf\u4e2a\u8bcd\u5143(token)\u3002ChatGPT\u901a\u8fc7\u4f7f\u7528\u5927\u91cf\u8bad\u7ec3\u6570\u636e\u6765\u6a21\u62df\u4eba\u7684\u8bed\u8a00\u884c\u4e3a\uff0c<\/span>\u751f\u6210\u4eba\u7c7b\u53ef\u4ee5\u7406\u89e3\u7684\u6587\u672c\uff0c<\/span>\u5e76\u6839\u636e\u4e0a\u4e0b\u6587\u8bed\u5883\u63d0\u4f9b\u6070\u5f53\u7684\u56de\u7b54\uff0c<\/span>\u751a\u81f3\u8fd8\u80fd\u505a\u53e5\u6cd5\u5206\u6790\u548c\u8bed\u4e49\u5206\u6790\uff0c<\/span>\u5e2e\u52a9\u7528\u6237\u8c03\u8bd5\u8ba1\u7b97\u673a\u7a0b\u5e8f\uff0c<\/span>\u5199\u8ba1\u7b97\u673a\u7a0b\u5e8f\u4ee3\u7801\uff0c<\/span>\u505a\u6570\u5b66\u9898\uff0c<\/span>\u800c\u4e14\u80fd\u591f\u901a\u8fc7\u4eba\u7c7b\u53cd\u9988\u7684\u4fe1\u606f\uff0c<\/span>\u4e0d\u65ad\u6539\u5584\u751f\u6210\u7684\u529f\u80fd,\u8fbe\u5230\u5f88\u5f3a\u7684\u81ea\u7136\u8bed\u8a00\u751f\u6210\u80fd\u529b\u3002ChatGPT\u7684\u8bad\u7ec3\u53c2\u6570\u8d8a\u6765\u8d8a\u591a\uff0c<\/span>\u6027\u80fd\u8d8a\u6765\u8d8a\u597d\u3002<\/span><\/p>\n
\n
2023\u5e743\u670817\u65e5, Open AI\u53d1\u5e03GPT-4\u3002GPT-4 \u5177\u6709\u5f3a\u5927\u7684\u8bc6\u56fe\u80fd\u529b\uff0c<\/span>\u6587\u5b57\u8f93\u5165\u9650\u5236\u75313\u5343\u8bcd\u63d0\u5347\u81f3 2.5 \u4e07\u8bcd,\u56de\u7b54\u95ee\u9898\u7684\u51c6\u786e\u5ea6\u663e\u8457\u63d0\u9ad8\uff0c<\/span>\u80fd\u591f\u751f\u6210\u6b4c\u8bcd\u3001\u521b\u610f\u6587\u672c,\u6539\u53d8\u6587\u672c\u7684\u5199\u4f5c\u98ce\u683c,\u751a\u81f3\u8fd8\u5177\u6709\u81ea\u52a8\u7ffb\u8bd1\u80fd\u529b\u30022023\u5e7411\u67087\u65e5,OpenAI\u4e3e\u884c\u5f00\u53d1\u65e5(DevDay),\u5ba3\u5e03\u4e86GPT-4\u7684\u4e00\u6b21\u5927\u5347\u7ea7,\u63a8\u51fa\u4e86GPT-4 Turbo,\u5f15\u8d77\u4e86\u5168\u4e16\u754c\u7684\u5bc6\u5207\u5173\u6ce8\u3002<\/span><\/p>\n
\n
\u6211\u4eec\u628a\u8fd9\u4e9b\u4e0d\u540c\u9636\u6bb5\u7684GPT\u7edf\u79f0\u4e3aGPTs\u7cfb\u5217\u3002GPTs\u7cfb\u5217\u7684\u6210\u529f\u5177\u6709\u5212\u65f6\u4ee3\u7684\u91cc\u7a0b\u7891\u610f\u4e49\uff0c\u662f\u5927\u8bed\u8a00\u6a21\u578b\u65f6\u4ee3\u6700\u4f1f\u5927\u7684\u6210\u679c\u4e4b\u4e00\uff0c\u8db3\u4ee5\u8f7d\u5165\u4eba\u5de5\u667a\u80fd\u53d1\u5c55\u53f2\u518c\u3002\u6211\u4eec\u5df2\u7ecf\u8fdb\u5165\u4e86\u5927\u8bed\u8a00\u6a21\u578b\u65f6\u4ee3(Era of Large Language Model)\u3002<\/span><\/p>\n
<\/section>\n
\n
\n
\n
\n
\n
\n
\n
03<\/strong><\/section>\n
02<\/strong><\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n
\u5927\u8bed\u8a00\u6a21\u578b\u65f6\u4ee3\u7684\u672f\u8bed\u7ffb\u8bd1<\/span><\/strong><\/span><\/section>\n<\/section>\n<\/section>\n
\n
\u672f\u8bed\u662f\u4eba\u7c7b\u7684\u79d1\u6280\u77e5\u8bc6\u5728\u81ea\u7136\u8bed\u8a00\u4e2d\u7684\u7ed3\u6676\u3002<\/span>\u5728\u5927\u8bed\u8a00\u6a21\u578b\u7684\u53d1\u5c55\u8fc7\u7a0b\u4e2d\uff0c\u51fa\u73b0\u4e86\u5927\u91cf\u8868\u793a\u65b0\u6280\u672f\u3001\u65b0\u6982\u5ff5\u7684\u672f\u8bed\u3002<\/span>\u8fd9\u4e9b\u672f\u8bed\u4e00\u822c\u9996\u5148\u4f7f\u7528\u82f1\u8bed\u8868\u8fbe\u3002<\/span>\u4e2d\u56fd\u4eba\u5728\u4f7f\u7528\u8fd9\u4e9b\u672f\u8bed\u7684\u65f6\u5019,\u6709\u5fc5\u8981\u628a\u8fd9\u4e9b\u65b0\u672f\u8bed\u7ffb\u8bd1\u6210\u6c49\u8bed\uff0c\u5b9e\u73b0\u672f\u8bed\u7684\u6c11\u65cf\u5316,\u4ee5\u4fbf\u5e7f\u5927\u4e2d\u56fd\u7528\u6237\u4f7f\u7528\u3002<\/span>2017\u5e74\u4ee5\u524d\uff0c\u5728\u6df1\u5ea6\u5b66\u4e60\u548c\u795e\u7ecf\u7f51\u7edc\u53d1\u5c55\u65f6\u671f\uff0c\u4e0d\u5c11\u82f1\u6587\u672f\u8bed\u5373\u88ab\u7ffb\u8bd1\u6210\u4e86\u6c49\u8bed(\u89c1\u88681)\u3002<\/span>\u8fd9\u6837\u7684\u672f\u8bed\u7ffb\u8bd1\uff0c\u4fbf\u4e8e\u4e2d\u56fd\u7528\u6237\u7406\u89e3\u548c\u4f7f\u7528,\u53d7\u5230\u6b22\u8fce\u3002<\/span><\/p>\n
\n
$\"\"$ <\/section>\n
\n
2017\u5e746\u6708,\u8c37\u6b4c\u516c\u53f8\u5728\u5176\u8bba\u6587\u300a\u6ce8\u610f\u529b\u5c31\u662f\u4f60\u4eec\u6240\u9700\u8981\u7684\u4e00\u5207\u300b(Attention Is All You Need)^{[\uff12<\/span><\/sup>^{]<\/span><\/sup>\u4e2d\u63d0\u51fa\u4e86\u4e00\u4e2a\u5b8c\u5168\u57fa\u4e8e\u6ce8\u610f\u529b\u673a\u5236\u7684\u9884\u8bad\u7ec3\u8bed\u8a00\u6a21\u578bTransformer,\u8be5\u6a21\u578b\u629b\u5f03\u4e86\u5728\u6b64\u4e4b\u524d\u7684\u5176\u4ed6\u91c7\u7528\u6ce8\u610f\u529b\u673a\u5236\u7684\u6a21\u578b\u6240\u4fdd\u7559\u7684\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7ed3\u6784\u4e0e\u5377\u79ef\u795e\u7ecf\u7f51\u7edc\u7ed3\u6784\uff0c\u6838\u5fc3\u90e8\u5206\u5b8c\u5168\u4f7f\u7528\u6ce8\u610f\u529b\u673a\u5236\u3002Transformer\u662f\u5b8c\u5168\u57fa\u4e8e\u6ce8\u610f\u529b\u673a\u5236\u7684\u6a21\u578b\uff0c\u5728\u5404\u9879\u4efb\u52a1\u7684\u5b8c\u6210\u548c\u6027\u80fd\u53d1\u6325\u65b9\u9762\u8868\u73b0\u4f18\u5f02\uff0c\u56e0\u6b64\u6210\u4e3a\u81ea\u7136\u8bed\u8a00\u7406\u89e3\u548c\u673a\u5668\u7ffb\u8bd1\u7684\u91cd\u8981\u57fa\u51c6\u6a21\u578b\u3002<\/span><\/p>\n}}
\n
\u7531\u4e8e\u5927\u8bed\u8a00\u6a21\u578b\u7684\u7814\u7a76\u548c\u53d1\u5c55\u6781\u5feb\uff0c\u5927\u91cf\u65b0\u51fa\u73b0\u7684\u672f\u8bed\u6765\u4e0d\u53ca\u7ffb\u8bd1\u6210\u6c49\u8bed\uff0c\u4e13\u4e1a\u9886\u57df\u5185\u90e8\u53ca\u4e13\u4e1a\u4eba\u5458\u4e4b\u95f4\uff0c\u7ecf\u5e38\u76f4\u63a5\u4f7f\u7528\u82f1\u8bed\u672f\u8bed\u3002Transformer\u8fd9\u4e2a\u65b0\u672f\u8bed\u4fbf\u662f\u5982\u6b64,\u5230\u76ee\u524d\u4e3a\u6b62\u6ca1\u6709\u786e\u5b9a\u7684\u5e7f\u4e3a\u8ba4\u53ef\u7684\u4e2d\u6587\u8bd1\u8bcd\u3002Transformer\u4e00\u8bcd\u5177\u6709\u591a\u91cd\u542b\u4e49:\u53d8\u538b\u5668\u3001\u53d8\u5f62\u91d1\u521a\u3001\u8f6c\u6362\u5668\u7b49,\u4f46\u662f\u4eba\u4eec\u89c9\u5f97\u8fd9\u4e9b\u90fd\u4e0d\u80fd\u8868\u8fbeTransformer\u5728\u4eba\u5de5\u667a\u80fd\u9886\u57df\u7684\u786e\u5207\u542b\u4e49\uff0c\u5f53\u4e0b\u53ea\u597d\u76f4\u63a5\u4f7f\u7528transformer\u8fd9\u4e2a\u82f1\u6587\u672f\u8bed^{[\uff13<\/span><\/sup>^{]<\/span><\/sup>\u3002<\/span><\/span><\/p>\n}}
\n
2019\u5e74,\u8c37\u6b4c\u516c\u53f8\u7814\u5236\u4e86Bidirectional Encoder Representations from Transformer\uff0c\u5373\u201c\u57fa\u4e8eTransformer\u7684\u53cc\u5411\u7f16\u7801\u5668\u8868\u793a\u201d\uff0c\u5230\u76ee\u524d\u4e3a\u6b62\u4e5f\u672a\u6709\u6070\u5f53\u7684\u4e2d\u6587\u7ffb\u8bd1\uff0c\u800c\u662f\u76f4\u63a5\u4f7f\u7528\u5176\u82f1\u6587\u7f29\u5199BERT\u3002<\/span><\/p>\n
\n
\u4e0e\u6b64\u540c\u65f6,\u5927\u8bed\u8a00\u6a21\u578b\u9886\u57df\u8fd8\u6709\u82e5\u5e72\u65b0\u51fa\u73b0\u7684\u4ea7\u54c1\u53ca\u6280\u672f\u672f\u8bed\uff0c\u90fd\u6ca1\u6709\u7ffb\u8bd1\u4e3a\u4e2d\u6587\uff0c\u4e1a\u5185\u666e\u904d\u4f7f\u7528\u5176\u82f1\u6587\u539f\u6587(\u89c1\u88682)\u3002\u5c31\u8fde\u6700\u8fd1\u516c\u5e03\u7684\u201c\u6c49\u8bed\u76d8\u70b92023\u201d\uff0c\u76f4\u63a5\u628a\u82f1\u6587\u672f\u8bedChatGPT\u4f5c\u4e3a\u6c49\u8bed\u70ed\u8bcd\u516c\u5e03\uff0c\u5c06\u5176\u79f0\u4e3a\u201c\u56fd\u9645\u8bcd(\u5982\u56fe2)\u3002<\/span><\/p>\n
\n
$\"\"$ <\/section>\n
<\/section>\n
$\"\"$ <\/section>\n
\n
\u4ee5\u4e0a\u60c5\u51b5\u8bf4\u660e\uff0c\u6211\u4eec\u5728\u63d0\u5021\u672f\u8bed\u6c11\u65cf\u5316\u7684\u540c\u65f6\uff0c\u672f\u8bed\u7684\u56fd\u9645\u5316\u4e5f\u6e10\u884c\u5176\u9053\uff0c\u800c\u4e14\u9010\u6e10\u6b65\u5165\u4e3b\u6d41\u3002<\/span><\/p>\n
\n
20\u4e16\u7eaa80\u5e74\u4ee3\uff0c\u5468\u6709\u5149\u5148\u751f\u5728\u201c\u6587\u5316\u4f20\u64ad\u548c\u672f\u8bed\u7ffb\u8bd1\u201d[4]\u4e00\u6587\u4e2d\u6307\u51fa\uff0c\u5728\u79d1\u5b66\u6280\u672f\u9886\u57df\uff0c\u53ef\u4ee5\u76f4\u63a5\u4f7f\u7528\u5916\u6587\u672f\u8bed\u800c\u4e0d\u7ffb\u8bd1\u6210\u6c49\u8bed\uff0c\u5b9e\u73b0\u672f\u8bed\u7684\u56fd\u9645\u5316\u3002\u4ed6\u4f3c\u4e4e\u9884\u89c1\u5230\u4e86\u4eca\u5929\u672f\u8bed\u56fd\u9645\u5316\u7684\u8fd9\u79cd\u5c40\u9762\uff0c\u6307\u51fa:\u201c\u5728\u9ad8\u5c42\u6b21\u7684\u4e13\u5bb6\u4e2d\uff0c\u672f\u8bed\u56fd\u9645\u5316\u4e0d\u4ec5\u662f\u53ef\u80fd\u7684,\u800c\u4e14\u662f\u5fc5\u8981\u7684\u3002\u5728\u4e00\u822c\u79d1\u6280\u5de5\u4f5c\u8005\u4e2d\u95f4\uff0c\u672f\u8bed\u6c11\u65cf\u5316\u6709\u5bb9\u6613\u5b66\u4e60\u7684\u597d\u5904\u3002\u201d \u201c\u4e2d\u56fd\u5982\u679c\u660e\u786e\u5730\u548c\u8ba4\u771f\u5730\u5b9e\u884c\u79d1\u6280\u53cc\u8bed\u8a00\u653f\u7b56\uff0c\u4e00\u65b9\u9762\u53ef\u4ee5\u4fdd\u6301\u672f\u8bed\u6c11\u65cf\u5316\u7684\u4f20\u7edf\uff0c\u4f7f\u5927\u4f17\u79d1\u6280\u5de5\u4f5c\u8005\u6bd4\u8f83\u5bb9\u6613\u5438\u6536\u79d1\u6280\u77e5\u8bc6\uff0c\u53e6\u4e00\u65b9\u9762\u53ef\u4ee5\u4e3a\u672f\u8bed\u56fd\u9645\u5316\u51c6\u5907\u5fc5\u8981\u6761\u4ef6\uff0c\u4f7f\u4e13\u4e1a\u79d1\u6280\u7814\u7a76\u8005\u8fce\u5934\u8d76\u4e0a\u8fc5\u731b\u53d1\u5c55\u7684\u4fe1\u606f\u5316\u65f6\u4ee3\u3002\u201d \u5468\u6709\u5149\u5148\u751f\u7684\u8fd9\u4e9b\u9884\u89c1,\u4eca\u5929\u6b63\u5728\u4e00\u6b65\u4e00\u6b65\u5730\u53d8\u6210\u73b0\u5b9e\u3002\u5728\u672f\u8bed\u5ba1\u5b9a\u5de5\u4f5c\u4e2d\uff0c\u6211\u4eec\u662f\u4e0d\u662f\u6709\u5fc5\u8981\u8ba4\u771f\u8003\u8651\u5468\u6709\u5149\u5148\u751f\u7684\u8fd9\u4e9b\u9884\u89c1\u5462?<\/span><\/p>\n
\n
\u5927\u8bed\u8a00\u6a21\u578b\u5bf9\u672f\u8bed\u5b66\u7814\u7a76\u63d0\u51fa\u4e86\u4e25\u5cfb\u7684\u6311\u6218\u3002\u8ba1\u7b97\u673a\u628a\u6587\u672c\u4e2d\u7684\u8bed\u8a00\u7b26\u53f7\u8f6c\u5316\u4e3a\u5411\u91cf,\u8bfb\u4e86\u4ebf\u4e07\u4e2a\u201c\u8bcd\u5143\u201d(token)\uff0c\u53ea\u8981\u6839\u636e\u4e0a\u4e0b\u6587\u5bf9\u4e8e\u4e0b\u4e00\u4e2a\u201c\u8bcd\u5143\u201d\u7684\u9884\u6d4b\u6765\u8c03\u6574\u53c2\u6570,\u5c31\u53ef\u4ee5\u751f\u6210\u7b26\u5408\u8bed\u6cd5\u4e14\u903b\u8f91\u901a\u987a\u7684\u6587\u672c\u3002\u53ef\u89c1\u6784\u6210\u65b0\u6587\u672c\u7684\u8109\u7edc\u5c31\u6f5c\u85cf\u5728\u5927\u89c4\u6a21\u7684\u6587\u672c\u6570\u636e\u4e4b\u4e2d\uff0c\u8fd9\u6837\u7684\u8109\u7edc\u7684\u5b9e\u8d28\u662f\u4ec0\u4e48\uff1f\u600e\u6837\u53d1\u73b0\u8fd9\u79cd\u8109\u7edc\uff1f\u600e\u6837\u4f7f\u7528\u8fd9\u79cd\u8109\u7edc\uff1f\u5176\u4e2d\u5305\u542b\u7740\u4f17\u591a\u7684\u672f\u8bed\u5b66\u95ee\u9898\uff0c\u5e94\u5f53\u5f15\u8d77\u6211\u4eec\u7684\u5bc6\u5207\u5173\u6ce8\u3002<\/span><\/p>\n
\n
\u5927\u8bed\u8a00\u6a21\u578b\u4e3a\u672f\u8bed\u5b66\u53d1\u5c55\u63d0\u4f9b\u4e86\u4e00\u4e2a\u5343\u8f7d\u96be\u9022\u7684\u597d\u673a\u4f1a\uff0c\u65b0\u65f6\u4ee3\u7684\u672f\u8bed\u5b66\u7814\u7a76\u5e94\u5f53\u9762\u5bf9\u8fd9\u4e9b\u95ee\u9898\uff0c\u4ece\u800c\u628a\u672f\u8bed\u5b66\u63a8\u8fdb\u5230\u4e00\u4e2a\u5d2d\u65b0\u7684\u9636\u6bb5\u3002<\/span><\/p>\n
\n
\n
\n
\n
\n
\n
\n
\n
\n
\n
\u4f5c\u8005\u7b80\u4ecb<\/section>\n<\/section>\n<\/section>\n<\/section>\n
\n
\n
\n
$\"\"$ <\/p>\n
\n
\u51af\u5fd7\u4f1f\uff081939\u2014\uff09\uff0c\u7537\uff0c\u8ba1\u7b97\u8bed\u8a00\u5b66\u5bb6\uff0c\u6559\u80b2\u90e8\u8bed\u8a00\u6587\u5b57\u5e94\u7528\u7814\u7a76\u6240\u7814\u7a76\u5458\uff0c\u4e2d\u56fd\u8ba1\u7b97\u673a\u5b66\u4f1a\u9ad8\u7ea7\u4f1a\u5458\u3002\u5317\u4eac\u5927\u5b66\u548c\u4e2d\u56fd\u79d1\u6280\u5927\u5b66\u7814\u7a76\u751f\u9662\u4e24\u6b21\u7814\u7a76\u751f\u6bd5\u4e1a\u300220\u4e16\u7eaa70\u5e74\u4ee3\u9996\u6b21\u6d4b\u51fa\u6c49\u5b57\u71b5\u4e3a9.65\u6bd4\u7279\uff0c\u4e3a\u6c49\u5b57\u591a\u516b\u4f4d\u7f16\u7801\u63d0\u4f9b\u4e86\u8bed\u8a00\u5b66\u652f\u6301\u3002\u7814\u5236\u4e86\u4e16\u754c\u4e0a\u7b2c\u4e00\u4e2a\u628a\u6c49\u8bed\u8bd1\u4e3a\u591a\u79cd\u5916\u8bed\u7684\u673a\u5668\u7ffb\u8bd1\u7cfb\u7edf\u3002\u957f\u671f\u4ece\u4e8b\u8bed\u8a00\u5b66\u548c\u8ba1\u7b97\u673a\u79d1\u5b66\u7684\u8de8\u5b66\u79d1\u7814\u7a76\uff0c\u51fa\u7248\u8bba\u845730\u591a\u90e8\uff0c\u53d1\u8868\u8bba\u6587400\u591a\u7bc7\uff0c\u4e3b\u6301\u7814\u5236\u56fd\u9645\u6807\u51c61\u9879\u3001\u56fd\u5bb6\u6807\u51c61\u9879\u3001\u56fd\u5bb6\u89c4\u83033\u9879\uff0c\u53c2\u4e0e\u7814\u5236\u56fd\u5bb6\u6807\u51c613\u9879\u3002\u66fe\u4efb\u4e2d\u56fd\u8bed\u6587\u73b0\u4ee3\u5316\u5b66\u4f1a\u526f\u4f1a\u957f\uff0c\u73b0\u4efb\u4e2d\u56fd\u4eba\u5de5\u667a\u80fd\u5b66\u4f1a\u7406\u4e8b\u3001\u300a\u4e2d\u6587\u4fe1\u606f\u5b66\u62a5\u300b\u987e\u95ee\u3001\u300a\u4e2d\u56fd\u79d1\u6280\u672f\u8bed\u300b\u526f\u4e3b\u7f16\u30022006\u5e74\u83b7\u5965\u5730\u5229\u7ef4\u65af\u7279\u5956\uff0c2018\u5e74\u83b7\u4e2d\u56fd\u8ba1\u7b97\u673a\u5b66\u4f1aNLPCC \u6770\u51fa\u8d21\u732e\u5956\uff0c2021\u5e74\u83b7\u4e2d\u56fd\u4e2d\u6587\u4fe1\u606f\u5b66\u4f1a\u4f1a\u58eb\u79f0\u53f7\uff0c2022\u5e74\u83b7\u9999\u6e2f\u5723\u5f17\u6717\u897f\u65af\u79d1\u6280\u4eba\u6587\u5956\u3002<\/section>\n
\u901a\u4fe1\u65b9\u5f0f\uff1a<\/section>\n
zwfengde2010@163.com<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n
\n
\n
\n<\/section>\n<\/section>\n
\n
\n
$\"\"$ <\/section>\n<\/section>\n
\n
END<\/strong><\/p>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n","protected":false},"excerpt":{"rendered":"
\u8bd1\u5b66\u6587\u732e | \u51af\u5fd7\u4f1f\uff1a\u5927\u8bed\u8a00\u6a21\u578b\u65f6\u4ee3\u7684\u672f\u8bed\u7ffb\u8bd1 \u7ffb\u8bd1\u6280\u672f\u6559\u80b2\u4e0e\u7814\u7a76\u00a02024\u5e7410\u670804\u65e5 00:00\u00a0\u9655\u897f […]<\/p>\n","protected":false},"author":1,"featured_media":34772,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[98],"tags":[],"class_list":["post-34769","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-98"],"_links":{"self":[{"href":"https:\/\/linguaresources.com\/index.php?rest_route=\/wp\/v2\/posts\/34769","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/linguaresources.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/linguaresources.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/linguaresources.com\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/linguaresources.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=34769"}],"version-history":[{"count":1,"href":"https:\/\/linguaresources.com\/index.php?rest_route=\/wp\/v2\/posts\/34769\/revisions"}],"predecessor-version":[{"id":34778,"href":"https:\/\/linguaresources.com\/index.php?rest_route=\/wp\/v2\/posts\/34769\/revisions\/34778"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/linguaresources.com\/index.php?rest_route=\/wp\/v2\/media\/34772"}],"wp:attachment":[{"href":"https:\/\/linguaresources.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=34769"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/linguaresources.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=34769"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/linguaresources.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=34769"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}