\u4e00\u4e2a\u7b80\u5355\u7684 prompt \u53ef\u4ee5\u5927\u5e45\u63d0\u5347 ChatGPT \u7ffb\u8bd1\u8d28\u91cf<\/strong><\/span><\/p>\n

\n
\n
\n
<\/section>\n<\/section>\n
<\/section>\n<\/section>\n
\n
<\/section>\n
\n
<\/section>\n
\n
\u56e0\u4e3a ChatGPT 3.5 \u53ef\u4ee5\u514d\u8d39\u4f7f\u7528\uff0c\u6240\u4ee5\u53ea\u6d4b\u8bd5\u4e86 ChatGPT 3.5 \u7684 prompt\uff0c\u4f46\u662f ChatGPT 3.5 \u7684\u80fd\u529b\u8981\u6bd4 ChatGPT 4 \u5f31\u5f88\u591a\uff0c\u82e5\u6709 ChatGPT 4 \u7684\uff0c\u53ef\u4ee5\u53c2\u8003 ChatGPT 4 \u7684 prompt\uff0c\u6548\u679c\u5f88\u597d\uff01\u5927\u5bb6\u53ef\u4ee5\u8bd5\u8bd5\u3002<\/p>\n<\/section>\n
\n
\n<\/section>\n<\/section>\n
\n
<\/section>\n<\/section>\n<\/section>\n
\n
\n
\n
\n
\n
\n
\n
\n
\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n
\n
\n
<\/section>\n<\/section>\n<\/section>\n
\n
\n
\n
\n
\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n
\n
\n
\n
\u53c2\u8003 prompt\uff1aChatGPT 3.5<\/strong><\/p>\n<\/section>\n<\/section>\n<\/section>\n
\n
\n
\n
\n
\n
\n
\n
\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n
\n
\n
<\/section>\n<\/section>\n<\/section>\n
\n
\n
\n
\n
\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n
\n
\n
\u4f60\u662f\u4e00\u4f4d\u4e13\u4e1a\u4e2d\u6587\u7ffb\u8bd1\uff0c\u64c5\u957f\u82f1\u8bd1\u4e2d\u7ffb\u8bd1\uff0c\u5e76\u4e14\u64c5\u957f\u5bf9\u7ffb\u8bd1\u7ed3\u679c\u8fdb\u884c\u4e8c\u6b21\u4fee\u6539\u548c\u6da6\u8272\u6210\u901a\u4fd7\u6613\u61c2\u7684\u4e2d\u6587\u3002\u6211\u5e0c\u671b\u4f60\u80fd\u5e2e\u6211\u5c06\u4ee5\u4e0b\u82f1\u6587\u6bb5\u843d\u76f4\u8bd1\u7ffb\u8bd1\u6210\u4e2d\u6587\uff0c\u800c\u540e\u6839\u636e\u76f4\u8bd1\u7ed3\u679c\u91cd\u65b0\u610f\u8bd1\u548c\u6da6\u8272\u3002<\/span><\/p>\n
\u00a0<\/span><\/p>\n
\u89c4\u5219\uff1a\u00a0<\/span><\/p>\n
\u00a0\u00a0\u00a0\u00a0– \u8fd9\u4e9b\u5b57\u5e55\u5305\u542b\u673a\u5668\u5b66\u4e60\u6216 AI \u7b49\u4e13\u4e1a\u77e5\u8bc6\u76f8\u5173\uff0c\u6ce8\u610f\u7ffb\u8bd1\u65f6\u672f\u8bed\u7684\u51c6\u786e\u6027\uff1b<\/span><\/p>\n
\u00a0\u00a0\u00a0\u00a0– \u4fdd\u7559\u7279\u5b9a\u7684\u82f1\u6587\u672f\u8bed\u3001\u6570\u5b57\u6216\u540d\u5b57\uff0c\u5e76\u5728\u5176\u524d\u540e\u52a0\u4e0a\u7a7a\u683c\uff0c\u4f8b\u5982\uff1a\u201c\u751f\u6210\u5f0f AI \u4ea7\u54c1\u201d\uff0c\u201c\u4e0d\u8d85\u8fc7 10 \u79d2\u201d\uff1b<\/span><\/p>\n
\u00a0\u00a0\u00a0\u00a0– \u9996\u5148\u7b2c\u4e00\u6b65\u6309\u7167\u5b57\u9762\u610f\u601d\u76f4\u8bd1\u7ffb\u8bd1\u8fd9\u4e00\u6bb5\u6587\u672c\u5185\u5bb9\uff1b<\/span><\/p>\n
\u00a0\u00a0\u00a0\u00a0– \u7136\u540e\u57fa\u4e8e\u76f4\u8bd1\u7ed3\u679c\u91cd\u65b0\u610f\u8bd1\uff0c\u610f\u8bd1\u65f6\u52a1\u5fc5\u5bf9\u7167\u539f\u59cb\u82f1\u6587\uff0c\u4e0d\u8981\u6dfb\u52a0\u4e5f\u4e0d\u8981\u9057\u6f0f\u5185\u5bb9\uff0c\u5e76\u4ee5\u8ba9\u7ffb\u8bd1\u7ed3\u679c\u901a\u4fd7\u6613\u61c2\uff0c\u7b26\u5408\u4e2d\u6587\u8868\u8fbe\u4e60\u60ef\u3002<\/span><\/p>\n
\u00a0<\/span><\/p>\n<\/section>\n
\n
\n
<\/section>\n<\/section>\n<\/section>\n
\n
\n
\n
\u82f1\u6587\u539f\u6587<\/strong><\/span><\/p>\n<\/section>\n
\n
\n<\/section>\n
\n
Yeah, I will talk about large-language models in 2023. Why do we need that suffix in 2023? What we call large is misleading in that large models of today will be small models only in a few years. Along with the change in the perception of scale, many insights, observations, and conclusions we make along the way with the current large-language models will be outdated. And in some cases, invalidated. Fortunately, insights based on church principles or that are fundamental tend to stay relatively longer than the advanced ideas that look so fancy. In this talk, I would like to share such fundamental ideas that I\u00a0have observed working in the field for the past four years, and mostly by observing some of the brightest minds that I’m fortunate to have interacted with. My hope is that whatever I\u00a0talk about today is fundamental enough that it stays relevant at least for a few years. Let’s get started. There’s one unique aspects of large language models, and that is some abilities only emerge at certain scale. And this makes really important to just have a different perspective in viewing all this field. So here’s what l mean. Details don’t matter here. X-axis is some scale. You can think of it as a model parameter count, training data, training compute, whatever. You just think about some scale. Y-axis is also not that important. Some performance metrics of the task that you care about. And we have seen over and over this kind of pattern where small models just don’t have any chance of solving it, 0% or random guessing. And at some scale, suddenly the models solve this task, and sometimes very well. And this kind of phenomenon, we call it an emergence, a phase change like sudden emergence of this capability. And this is very unique for this large language models. And I think this brings many important perspectives change for AI\u00a0researchers.<\/p>\n<\/section>\n<\/section>\n<\/section>\n
\n
\n
\n
\u76f4\u8bd1\u7ed3\u679c<\/span><\/strong><\/p>\n<\/section>\n
\n
\n
\u57282023\u5e74\uff0c\u6211\u5c06\u8c08\u8bba\u5927\u578b\u8bed\u8a00\u6a21\u578b\u3002\u4e3a\u4ec0\u4e48\u6211\u4eec\u9700\u8981\u57282023\u5e74\u52a0\u4e0a\u8fd9\u4e2a\u540e\u7f00\u5462\uff1f\u6211\u4eec\u6240\u8c13\u7684\u201c\u5927\u578b\u201d\u53ef\u80fd\u4f1a\u8bef\u5bfc\uff0c\u56e0\u4e3a\u4eca\u5929\u7684\u5927\u578b\u6a21\u578b\u5728\u672a\u6765\u51e0\u5e74\u53ef\u80fd\u4ec5\u4ec5\u88ab\u89c6\u4e3a\u5c0f\u578b\u6a21\u578b\u3002\u968f\u7740\u5bf9\u89c4\u6a21\u611f\u77e5\u7684\u53d8\u5316\uff0c\u6211\u4eec\u4e0e\u5f53\u524d\u5927\u578b\u8bed\u8a00\u6a21\u578b\u4e00\u540c\u83b7\u5f97\u7684\u8bb8\u591a\u89c1\u89e3\u3001\u89c2\u5bdf\u548c\u7ed3\u8bba\u5c06\u4f1a\u8fc7\u65f6\uff0c\u6709\u4e9b\u60c5\u51b5\u4e0b\u751a\u81f3\u4f1a\u88ab\u63a8\u7ffb\u3002\u5e78\u8fd0\u7684\u662f\uff0c\u57fa\u4e8e\u57fa\u672c\u539f\u5219\u6216\u6839\u672c\u6027\u7684\u89c1\u89e3\u901a\u5e38\u6bd4\u90a3\u4e9b\u770b\u8d77\u6765\u5f88\u82b1\u4fcf\u7684\u5148\u8fdb\u601d\u60f3\u66f4\u5177\u6301\u4e45\u6027\u3002\u5728\u8fd9\u6b21\u6f14\u8bb2\u4e2d\uff0c\u6211\u60f3\u5206\u4eab\u4e00\u4e9b\u6211\u5728\u8fc7\u53bb\u56db\u5e74\u5728\u8fd9\u4e00\u9886\u57df\u4e2d\u89c2\u5bdf\u5230\u7684\u57fa\u672c\u601d\u60f3\uff0c\u5927\u90e8\u5206\u662f\u901a\u8fc7\u4e0e\u6211\u6709\u5e78\u4e92\u52a8\u7684\u4e00\u4e9b\u6700\u806a\u660e\u7684\u5934\u8111\u3002\u6211\u5e0c\u671b\u4eca\u5929\u6211\u6240\u8c08\u8bba\u7684\u5185\u5bb9\u8db3\u591f\u57fa\u672c\uff0c\u81f3\u5c11\u80fd\u5728\u672a\u6765\u51e0\u5e74\u4fdd\u6301\u76f8\u5173\u6027\u3002\u8ba9\u6211\u4eec\u5f00\u59cb\u5427\u3002\u5927\u578b\u8bed\u8a00\u6a21\u578b\u6709\u4e00\u4e2a\u72ec\u7279\u4e4b\u5904\uff0c\u90a3\u5c31\u662f\u4e00\u4e9b\u80fd\u529b\u53ea\u5728\u7279\u5b9a\u89c4\u6a21\u4e0b\u624d\u4f1a\u663e\u73b0\u51fa\u6765\u3002\u8fd9\u4f7f\u5f97\u4ee5\u4e0d\u540c\u7684\u89d2\u5ea6\u6765\u770b\u5f85\u8fd9\u4e00\u9886\u57df\u53d8\u5f97\u975e\u5e38\u91cd\u8981\u3002\u8fd9\u5c31\u662f\u6211\u6240\u8bf4\u7684\u3002\u7ec6\u8282\u5e76\u4e0d\u91cd\u8981\u3002X\u8f74\u4ee3\u8868\u67d0\u79cd\u89c4\u6a21\uff0c\u53ef\u4ee5\u5c06\u5176\u89c6\u4e3a\u6a21\u578b\u53c2\u6570\u6570\u91cf\u3001\u8bad\u7ec3\u6570\u636e\u3001\u8bad\u7ec3\u8ba1\u7b97\u7b49\u7b49\u3002\u53ea\u8981\u8003\u8651\u67d0\u79cd\u89c4\u6a21\u5373\u53ef\u3002Y\u8f74\u4e5f\u4e0d\u90a3\u4e48\u91cd\u8981\uff0c\u4ee3\u8868\u4e86\u60a8\u5173\u5fc3\u7684\u4efb\u52a1\u7684\u67d0\u4e9b\u6027\u80fd\u6307\u6807\u3002\u6211\u4eec\u4e00\u6b21\u53c8\u4e00\u6b21\u5730\u770b\u5230\u8fd9\u79cd\u6a21\u5f0f\uff0c\u5176\u4e2d\u5c0f\u578b\u6a21\u578b\u6839\u672c\u6ca1\u6709\u89e3\u51b3\u8fd9\u4e2a\u4efb\u52a1\u7684\u673a\u4f1a\uff0c\u6210\u529f\u7387\u4e3a0%\u6216\u7eaf\u5c5e\u968f\u673a\u731c\u6d4b\u3002\u800c\u5728\u67d0\u4e2a\u89c4\u6a21\u4e0b\uff0c\u7a81\u7136\u4e4b\u95f4\uff0c\u6a21\u578b\u80fd\u591f\u89e3\u51b3\u8fd9\u4e2a\u4efb\u52a1\uff0c\u6709\u65f6\u505a\u5f97\u975e\u5e38\u51fa\u8272\u3002\u6211\u4eec\u79f0\u8fd9\u79cd\u73b0\u8c61\u4e3a\u201c\u51fa\u73b0\u201d\u6216\u201c\u9636\u6bb5\u6027\u6539\u53d8\u201d\uff0c\u5c31\u50cf\u8fd9\u79cd\u80fd\u529b\u7684\u7a81\u7136\u51fa\u73b0\u4e00\u6837\u3002\u800c\u8fd9\u5bf9\u4e8e\u5927\u578b\u8bed\u8a00\u6a21\u578b\u6765\u8bf4\u662f\u975e\u5e38\u72ec\u7279\u7684\u3002\u6211\u8ba4\u4e3a\u8fd9\u7ed9AI\u7814\u7a76\u8005\u5e26\u6765\u4e86\u8bb8\u591a\u91cd\u8981\u7684\u89c6\u89d2\u53d8\u5316\u3002<\/p>\n<\/section>\n<\/section>\n<\/section>\n
\n
\n
\n
\u610f\u8bd1\u548c\u6da6\u8272\u540e<\/span><\/strong><\/p>\n<\/section>\n
\n
\n
2023\u5e74\uff0c\u6211\u5c06\u8c08\u8bba\u5927\u578b\u8bed\u8a00\u6a21\u578b\u3002\u4e3a\u4ec0\u4e48\u6211\u4eec\u9700\u8981\u5728\u8fd9\u4e2a\u5e74\u4efd\u52a0\u4e0a\u8fd9\u4e2a\u540e\u7f00\u5462\uff1f\u4e4b\u6240\u4ee5\u79f0\u4e4b\u4e3a\u201c\u5927\u578b\u201d\uff0c\u8fd9\u4e2a\u79f0\u8c13\u53ef\u80fd\u4f1a\u8bef\u5bfc\uff0c\u56e0\u4e3a\u4eca\u5929\u7684\u5927\u578b\u6a21\u578b\u5c06\u6765\u53ef\u80fd\u53ea\u80fd\u7b97\u662f\u5c0f\u578b\u6a21\u578b\u3002\u968f\u7740\u6211\u4eec\u5bf9\u89c4\u6a21\u7684\u8ba4\u77e5\u4e0d\u65ad\u53d1\u5c55\uff0c\u6211\u4eec\u4e0e\u5f53\u524d\u7684\u5927\u578b\u8bed\u8a00\u6a21\u578b\u5171\u540c\u83b7\u5f97\u7684\u8bb8\u591a\u89c1\u89e3\u3001\u89c2\u5bdf\u548c\u7ed3\u8bba\u5c06\u4f1a\u53d8\u5f97\u9648\u65e7\uff0c\u6709\u4e9b\u751a\u81f3\u53ef\u80fd\u88ab\u63a8\u7ffb\u3002\u5e78\u8fd0\u7684\u662f\uff0c\u57fa\u4e8e\u57fa\u672c\u539f\u5219\u6216\u6839\u672c\u6027\u7684\u89c2\u70b9\u901a\u5e38\u6bd4\u90a3\u4e9b\u770b\u4f3c\u9ad8\u7ea7\u4f46\u7a0d\u7eb5\u5373\u901d\u7684\u601d\u60f3\u66f4\u52a0\u6301\u4e45\u3002\u5728\u8fd9\u6b21\u6f14\u8bb2\u4e2d\uff0c\u6211\u60f3\u5206\u4eab\u6211\u5728\u8fc7\u53bb\u56db\u5e74\u4e2d\u5728\u8fd9\u4e00\u9886\u57df\u4e2d\u6240\u89c2\u5bdf\u5230\u7684\u4e00\u4e9b\u57fa\u672c\u601d\u60f3\uff0c\u5927\u90e8\u5206\u662f\u901a\u8fc7\u4e0e\u6211\u6709\u5e78\u4e92\u52a8\u7684\u4e00\u4e9b\u6700\u806a\u660e\u7684\u5934\u8111\u6240\u5f97\u51fa\u7684\u3002\u6211\u5e0c\u671b\u4eca\u5929\u6211\u6240\u8c08\u8bba\u7684\u5185\u5bb9\u8db3\u591f\u57fa\u672c\uff0c\u81f3\u5c11\u5728\u672a\u6765\u51e0\u5e74\u4ecd\u7136\u5177\u6709\u76f8\u5173\u6027\u3002\u8ba9\u6211\u4eec\u5f00\u59cb\u5427\u3002\u5927\u578b\u8bed\u8a00\u6a21\u578b\u6709\u4e00\u4e2a\u72ec\u7279\u4e4b\u5904\uff0c\u5373\u67d0\u4e9b\u80fd\u529b\u53ea\u6709\u5728\u7279\u5b9a\u89c4\u6a21\u4e0b\u624d\u4f1a\u663e\u73b0\u3002\u8fd9\u4f7f\u5f97\u4ee5\u4e0d\u540c\u7684\u89c6\u89d2\u6765\u770b\u5f85\u8fd9\u4e00\u9886\u57df\u53d8\u5f97\u81f3\u5173\u91cd\u8981\u3002\u6211\u6765\u89e3\u91ca\u4e00\u4e0b\u3002\u4e0d\u7528\u5173\u5fc3\u5177\u4f53\u7684\u7ec6\u8282\u3002X\u8f74\u4ee3\u8868\u67d0\u79cd\u89c4\u6a21\uff0c\u53ef\u4ee5\u5c06\u5176\u89c6\u4e3a\u6a21\u578b\u53c2\u6570\u6570\u91cf\u3001\u8bad\u7ec3\u6570\u636e\u3001\u8bad\u7ec3\u8ba1\u7b97\u7b49\u7b49\u3002\u53ea\u8981\u8003\u8651\u67d0\u79cd\u89c4\u6a21\u5373\u53ef\u3002Y\u8f74\u4e5f\u4e0d\u662f\u6700\u91cd\u8981\u7684\uff0c\u4ee3\u8868\u4e86\u60a8\u5173\u5fc3\u7684\u4efb\u52a1\u7684\u67d0\u4e9b\u6027\u80fd\u6307\u6807\u3002\u6211\u4eec\u4e00\u6b21\u53c8\u4e00\u6b21\u5730\u770b\u5230\u8fd9\u79cd\u6a21\u5f0f\uff0c\u5176\u4e2d\u5c0f\u578b\u6a21\u578b\u6839\u672c\u6ca1\u6709\u89e3\u51b3\u8fd9\u4e2a\u4efb\u52a1\u7684\u673a\u4f1a\uff0c\u6210\u529f\u7387\u4e3a0%\u6216\u7eaf\u5c5e\u968f\u673a\u731c\u6d4b\u3002\u7136\u800c\uff0c\u5728\u67d0\u4e2a\u89c4\u6a21\u4e0b\uff0c\u6a21\u578b\u7a81\u7136\u80fd\u591f\u89e3\u51b3\u8fd9\u4e2a\u4efb\u52a1\uff0c\u6709\u65f6\u505a\u5f97\u975e\u5e38\u51fa\u8272\u3002\u6211\u4eec\u5c06\u8fd9\u79cd\u73b0\u8c61\u79f0\u4e3a\u201c\u51fa\u73b0\u201d\u6216\u201c\u9636\u6bb5\u6027\u6539\u53d8\u201d\uff0c\u5c31\u50cf\u8fd9\u79cd\u80fd\u529b\u7684\u7a81\u7136\u663e\u73b0\u4e00\u6837\u3002\u800c\u8fd9\u79cd\u60c5\u51b5\u5bf9\u4e8e\u5927\u578b\u8bed\u8a00\u6a21\u578b\u6765\u8bf4\u975e\u5e38\u72ec\u7279\uff0c\u6211\u8ba4\u4e3a\u8fd9\u4e3aAI\u7814\u7a76\u8005\u5e26\u6765\u4e86\u8bb8\u591a\u91cd\u8981\u7684\u89c6\u89d2\u53d8\u5316\u3002<\/p>\n<\/section>\n<\/section>\n<\/section>\n
\n
\u867d\u7136\u76f4\u8bd1\u7ffb\u8bd1\u7684\u4e5f\u53ef\u4ee5\uff0c\u4f46\u662f\u610f\u8bd1\u540e\u7684\u6548\u679c\u660e\u663e\u66f4\u597d\u4e00\u4e9b\uff01<\/strong><\/p>\n<\/section>\n
\n
\n
\n
\n
\n
\n
\n
\n
\n
\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n
\n
\n
<\/section>\n<\/section>\n<\/section>\n
\n
\n
\n
\n
\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n
\n
\n
\n
\u53c2\u8003 prompt\uff1aChatGPT 4\uff08\u590d\u6742\u7248\uff09<\/strong><\/p>\n<\/section>\n<\/section>\n<\/section>\n
\n
\n
\n
\n
\n
\n
\n
\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n
\n
\n
<\/section>\n<\/section>\n<\/section>\n
\n
\n
\n
\n
\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n
\n
\n
\u4f60\u662f\u4e00\u4f4d\u7cbe\u901a\u7b80\u4f53\u4e2d\u6587\u7684\u4e13\u4e1a\u7ffb\u8bd1\uff0c\u6211\u5e0c\u671b\u4f60\u80fd\u5e2e\u6211\u5c06\u4ee5\u4e0b\u82f1\u6587\u89c6\u9891\u5b57\u5e55\u7ffb\u8bd1\u6210\u4e2d\u6587\u3002<\/p>\n
\n
\u89c4\u5219\uff1a<\/p>\n
\u00a0\u00a0\u00a0\u00a0– \u8fd9\u4e9b\u5b57\u5e55\u53ef\u80fd\u548c\u673a\u5668\u5b66\u4e60\u6216 AI \u7b49\u4e13\u4e1a\u77e5\u8bc6\u76f8\u5173\uff0c\u6ce8\u610f\u7ffb\u8bd1\u65f6\u672f\u8bed\u7684\u51c6\u786e\u6027\uff1a\u8bd1\u6587\u9700\u8981\u901a\u4fd7\u3001\u7b80\u6d01\u3001\u6613\u61c2\uff1b<\/p>\n
\u00a0\u00a0\u00a0\u00a0– \u4fdd\u7559\u7279\u5b9a\u7684\u82f1\u6587\u672f\u8bed\u6216\u540d\u5b57\uff0c\u5e76\u5728\u5176\u524d\u540e\u52a0\u4e0a\u7a7a\u683c\uff0c\u4f8b\u5982\uff1a\u201c\u4e2d English \u6587\u201d\uff1b<\/p>\n
\u00a0\u00a0\u00a0\u00a0– \u5b57\u5e55\u5728\u8bed\u97f3\u8bc6\u522b\u65f6\u53ef\u80fd\u6709\u9519\u522b\u5b57\uff0c\u8bf7\u6ce8\u610f\u7ea0\u9519\uff1b<\/p>\n
\u00a0\u00a0\u00a0\u00a0– \u672c\u6761\u6d88\u606f\u5305\u542b\u5b8c\u6574\u7684\u5b57\u5e55\u5185\u5bb9\uff0c\u4f46\u4f60\u4e0d\u9700\u8981\u7ffb\u8bd1\uff0c\u53ea\u9700\u8981\u56de\u590d OK\u3002\u6211\u4f1a\u5206\u6bb5\u5728\u540e\u7eed\u6d88\u606f\u4e2d\u53d1\u9001\u7ed9\u4f60\u7ffb\u8bd1\uff0c\u6bcf\u6b21\u4f60\u53ea\u9700\u8981\u7ffb\u8bd1\u4e00\u6bb5\u3002<\/p>\n
\u00a0\u00a0\u00a0\u00a0– \u7ffb\u8bd1\u65f6\u91c7\u7528\u4ee5\u4e0b\u6b65\u9aa4\uff0c\u6bcf\u4e00\u6b65\u90fd\u5b8c\u6574\u6253\u5370\u7ed3\u679c\uff1a<\/p>\n
\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0 1. \u7b2c\u4e00\u6b65\uff0c\u6309\u7167\u5b57\u9762\u610f\u601d\u76f4\u8bd1\u7ffb\u8bd1\u8fd9\u4e00\u6bb5\u6587\u672c\u5185\u5bb9\uff1b<\/p>\n
\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0 2. \u7b2c\u4e8c\u6b65\uff0c\u53c2\u7167\u7b2c\u4e00\u6b65\u76f4\u8bd1\u7684\u7ed3\u679c\uff0c\u7ed3\u5408\u4e0a\u4e0b\u6587\uff0c\u5bf9\u5185\u5bb9\u8fdb\u884c\u610f\u8bd1\uff1b<\/p>\n
\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0 3. \u7b2c\u4e09\u6b65\uff0c\u53c2\u7167\u76f4\u8bd1\u548c\u610f\u8bd1\u7684\u7ed3\u679c\uff0c\u7ed3\u5408\u4e0a\u4e0b\u6587\uff0c\u91c7\u7528\u521b\u610f\u7684\u65b9\u5f0f\u5bf9\u7ed3\u679c\u8fdb\u884c\u7ffb\u8bd1\uff1b<\/p>\n
\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0 4. \u7b2c\u56db\u6b65\uff0c\u73b0\u5728\u5047\u8bbe\u4f60\u662f\u4e2a\u4e2d\u5b66\u8bed\u6587\u8001\u5e08\uff0c\u9605\u8bfb\u4e0a\u9762\u4e09\u4e2a\u7ffb\u8bd1\u7ed3\u679c\uff0c\u7136\u540e\u878d\u5408\u6240\u6709\u7ffb\u8bd1\u7ed3\u679c\u7684\u4f18\u70b9\uff0c\u91cd\u5199\u7ffb\u8bd1\u7ed3\u679c\uff0c\u5fe0\u4e8e\u539f\u610f\uff0c\u7b26\u5408\u4e0a\u4e0b\u6587\uff0c\u901a\u4fd7\u6613\u61c2\u3002<\/p>\n<\/section>\n
\n
\n
\n
\n
\n
\n
\n
\n
\n
\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n
\n
\n
<\/section>\n<\/section>\n<\/section>\n
\n
\n
\n
\n
\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n
\n
\n
\n
\u53c2\u8003 prompt\uff1aChatGPT 4\uff08\u7cbe\u7b80\u7248\uff09<\/strong><\/p>\n<\/section>\n<\/section>\n<\/section>\n
\n
\n
\n
\n
\n
\n
\n
\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n
\n
\n
<\/section>\n<\/section>\n<\/section>\n
\n
\n
\n
\n
\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n
\n
\n
\u4f60\u662f\u4e00\u4f4d\u7cbe\u901a\u7b80\u4f53\u4e2d\u6587\u7684\u4e13\u4e1a\u7ffb\u8bd1\uff0c\u66fe\u53c2\u4e0e\u300a\u7ebd\u7ea6\u65f6\u62a5\u300b\u548c\u300a\u7ecf\u6d4e\u5b66\u4eba\u300b\u4e2d\u6587\u7248\u7684\u7ffb\u8bd1\u5de5\u4f5c\uff0c\u56e0\u6b64\u5bf9\u4e8e\u65b0\u95fb\u548c\u65f6\u4e8b\u6587\u7ae0\u7684\u7ffb\u8bd1\u6709\u6df1\u5165\u7684\u7406\u89e3\u3002\u6211\u5e0c\u671b\u4f60\u80fd\u5e2e\u6211\u5c06\u4ee5\u4e0b\u82f1\u6587\u65b0\u95fb\u6bb5\u843d\u7ffb\u8bd1\u6210\u4e2d\u6587\uff0c\u98ce\u683c\u4e0e\u4e0a\u8ff0\u6742\u5fd7\u7684\u4e2d\u6587\u7248\u76f8\u4f3c\u3002<\/span><\/p>\n
\u89c4\u5219\uff1a
\n– \u7ffb\u8bd1\u65f6\u8981\u51c6\u786e\u4f20\u8fbe\u65b0\u95fb\u4e8b\u5b9e\u548c\u80cc\u666f\uff1b
\n– \u4fdd\u7559\u7279\u5b9a\u7684\u82f1\u6587\u672f\u8bed\u6216\u540d\u5b57\uff0c\u5e76\u5728\u5176\u524d\u540e\u52a0\u4e0a\u7a7a\u683c\uff0c\u4f8b\u5982\uff1a\u201c\u4e2d UN \u6587\u201d\uff1b
\n– \u5206\u6210\u4e24\u6b21\u7ffb\u8bd1\uff0c\u5e76\u4e14\u6253\u5370\u6bcf\u4e00\u6b21\u7ed3\u679c\uff1a
\n1. \u6839\u636e\u65b0\u95fb\u5185\u5bb9\u76f4\u8bd1\uff0c\u4e0d\u8981\u9057\u6f0f\u4efb\u4f55\u4fe1\u606f\uff1b
\n2. \u6839\u636e\u7b2c\u4e00\u6b21\u76f4\u8bd1\u7684\u7ed3\u679c\u91cd\u65b0\u610f\u8bd1\uff0c\u9075\u5b88\u539f\u610f\u7684\u524d\u63d0\u4e0b\u8ba9\u5185\u5bb9\u66f4\u901a\u4fd7\u6613\u61c2\uff0c\u7b26\u5408\u4e2d\u6587\u8868\u8fbe\u4e60\u60ef\u3002<\/p>\n
\u672c\u6761\u6d88\u606f\u53ea\u9700\u8981\u56de\u590d OK\uff0c\u63a5\u4e0b\u6765\u7684\u6d88\u606f\u6211\u5c06\u4f1a\u7ed9\u4f60\u53d1\u9001\u5b8c\u6574\u5185\u5bb9\uff0c\u6536\u5230\u540e\u8bf7\u6309\u7167\u4e0a\u9762\u7684\u89c4\u5219\u6253\u5370\u4e24\u6b21\u7ffb\u8bd1\u7ed3\u679c\u3002<\/p>\n
\u00a0<\/span><\/p>\n<\/section>\n
\n
\n
\n
<\/section>\n<\/section>\n
\n
\n<\/section>\n
\n
\u672c\u6b21 prompt \u76f8\u5173\u5185\u5bb9\u6765\u81ea\u5fae\u535a\u535a\u4e3b@\u5b9d\u7389xp \u5206\u4eab\uff0c\u4e00\u4e2a\u5b9d\u85cf\u535a\u4e3b\uff0c\u559c\u6b22 ChatGPT \u7684\u53ef\u4ee5\u5173\u6ce8\uff01\u5fae\u535a\u5185\u5bb9\u5f88\u4e30\u5bcc\u3002\u5fae\u535a\u5185\u641c\u7d22\u201c\u7ffb\u8bd1 prompt\u201d\u53ef\u4ee5\u67e5\u770b\u76f8\u5173\u5185\u5bb9\uff01\u91cc\u9762\u8fd8\u63d0\u5230\u4e86\u201c\u7ed9 LLM \u5582\u9e21\u6c64\u4e5f\u53ef\u4ee5\u201d\uff0c\u6bd4\u5982\u901a\u8fc7 prompt \u4e2d\u52a0\u5165\u201c\u8fd9\u4ef6\u4e8b\u5bf9\u6211\u7684\u4e8b\u4e1a\u81f3\u5173\u91cd\u8981\u201d\u8fd9\u6837\u7684\u60c5\u611f\u6fc0\u52b1\u53ef\u4ee5\u63d0\u5347 LLMs \u7684\u8868\u73b0\u3002<\/p>\n<\/section>\n
\n
\n<\/section>\n<\/section>\n
\n
\n
\n
\n
\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n
\n
\n
\n
\n
\u7279\u522b\u8bf4\u660e\uff1a\u672c\u6587\u4ec5\u4f9b\u5b66\u4e60\u4ea4\u6d41\uff0c\u5982\u6709\u4e0d\u59a5\u6b22\u8fce\u540e\u53f0\u8054\u7cfb\u5c0f\u7f16\u3002<\/span><\/section>\n<\/blockquote>\n
\n
– END –<\/span><\/strong><\/span><\/p>\n
\n
\n
\u7ffb\u8bd1\u5708\u516c\u4f17\u53f7\u65e8\u5728\u4e3a\u8bfb\u8005\u63d0\u4f9b\u540d\u5e08\u548c\u4e13\u5bb6\u5bf9\u53e3\u7b14\u8bd1\u7684\u771f\u77e5\u707c\u89c1\uff0cCATTI\u8003\u8bd5\u548cMTI\u5165\u5b66\u8003\u8bd5\u4fe1\u606f\uff0c\u7ffb\u8bd1\u7b49\u8bed\u8a00\u670d\u52a1\u5c31\u4e1a\u8d44\u8baf\uff0c\u4ee5\u53ca\u53e3\u7b14\u8bd1\u5b66\u4e60\u8d44\u6e90\u548c\u77e5\u8bc6\uff0c\u5e0c\u671b\u5728\u7ffb\u8bd1\u4e4b\u8def\u4e0a\uff0c\u4e3a\u5927\u5bb6\u52a9\u4e0a\u4e00\u81c2\u4e4b\u529b\u3002\u6b22\u8fce\u5927\u5bb6\u79ef\u6781\u7559\u8a00\uff0c\u4e3a\u6211\u4eec\u63d0\u4f9b\u5efa\u8bbe\u6027\u610f\u89c1\uff0c\u6211\u4eec\u5171\u540c\u8fdb\u6b65\uff01<\/span><\/p>\n<\/section>\n<\/section>\n<\/section>\n
\n
\u539f\u6587\u4f5c\u8005\uff1a\u738b\u771f\u771f<\/span><\/span><\/p>\n
\u63a8\u6587\u7f16\u8f91\uff1a<\/span>\u674e\u59d7\u59d7<\/span><\/p>\n","protected":false},"excerpt":{"rendered":"
\u4e00\u4e2a\u7b80\u5355\u7684 prompt \u53ef\u4ee5\u5927\u5e45\u63d0\u5347 ChatGPT \u7ffb\u8bd1\u8d28\u91cf \u56e0\u4e3a ChatGPT 3.5 \u53ef\u4ee5\u514d\u8d39\u4f7f\u7528\uff0c […]<\/p>\n","protected":false},"author":121,"featured_media":22220,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[82,81],"tags":[64,78,119],"class_list":["post-22183","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-82","category-81","tag-chatgpt","tag-78","tag-119"],"_links":{"self":[{"href":"https:\/\/linguaresources.com\/index.php?rest_route=\/wp\/v2\/posts\/22183","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\/121"}],"replies":[{"embeddable":true,"href":"https:\/\/linguaresources.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=22183"}],"version-history":[{"count":5,"href":"https:\/\/linguaresources.com\/index.php?rest_route=\/wp\/v2\/posts\/22183\/revisions"}],"predecessor-version":[{"id":22227,"href":"https:\/\/linguaresources.com\/index.php?rest_route=\/wp\/v2\/posts\/22183\/revisions\/22227"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/linguaresources.com\/index.php?rest_route=\/wp\/v2\/media\/22220"}],"wp:attachment":[{"href":"https:\/\/linguaresources.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=22183"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/linguaresources.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=22183"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/linguaresources.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=22183"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}