{"id":1342,"date":"2025-06-10T20:31:00","date_gmt":"2025-06-10T12:31:00","guid":{"rendered":"http:\/\/www.applicationsofllm.com\/?p=1342"},"modified":"2025-09-17T20:33:48","modified_gmt":"2025-09-17T12:33:48","slug":"%e5%94%af%e5%bf%ab%e4%b8%8d%e7%a0%b4%ef%bc%9a%e4%b8%8a%e6%b5%b7ai-lab-82%e9%a1%b5%e7%bb%bc%e8%bf%b0%e5%b8%a6%e4%bd%a0%e6%84%9f%e5%8f%97llm%e9%ab%98%e6%95%88%e6%9e%b6%e6%9e%84%e7%9a%84%e9%ad%85","status":"publish","type":"post","link":"http:\/\/www.applicationsofllm.com\/index.php\/2025\/06\/10\/%e5%94%af%e5%bf%ab%e4%b8%8d%e7%a0%b4%ef%bc%9a%e4%b8%8a%e6%b5%b7ai-lab-82%e9%a1%b5%e7%bb%bc%e8%bf%b0%e5%b8%a6%e4%bd%a0%e6%84%9f%e5%8f%97llm%e9%ab%98%e6%95%88%e6%9e%b6%e6%9e%84%e7%9a%84%e9%ad%85\/","title":{"rendered":"\u552f\u5feb\u4e0d\u7834\uff1a\u4e0a\u6d77AI Lab 82\u9875\u7efc\u8ff0\u5e26\u4f60\u611f\u53d7LLM\u9ad8\u6548\u67b6\u6784\u7684\u9b45\u529b"},"content":{"rendered":"\n<p><\/p>\n\n\n\n<p><sup>\u00a0\u56fe 1\uff1a\u5e38\u89c1\u957f\u5e8f\u5217\u573a\u666f<\/sup><\/p>\n\n\n\n<p>\u8fd1\u5e74\u6765\uff0c\u5927\u8bed\u8a00\u6a21\u578b\uff08LLMs\uff09\u5c55\u73b0\u51fa\u5f3a\u5927\u7684\u8bed\u8a00\u7406\u89e3\u4e0e\u751f\u6210\u80fd\u529b\uff0c\u63a8\u52a8\u4e86\u6587\u672c\u751f\u6210\u3001\u4ee3\u7801\u751f\u6210\u3001\u95ee\u7b54\u3001\u7ffb\u8bd1\u7b49\u4efb\u52a1\u7684\u7a81\u7834\u3002\u4ee3\u8868\u6027\u6a21\u578b\u5982 GPT\u3001Claude\u3001Gemini\u3001DeepSeek\u3001Qwen \u7b49\uff0c\u5df2\u7ecf\u6df1\u523b\u6539\u53d8\u4e86\u4eba\u673a\u4ea4\u4e92\u65b9\u5f0f\u3002LLMs \u7684\u8fb9\u754c\u4e5f\u4e0d\u6b62\u4e8e\u8bed\u8a00\u548c\u7b80\u5355\u95ee\u7b54\u3002\u968f\u7740\u591a\u6a21\u6001\uff08VLMs\uff09\u4e0e\u63a8\u7406\u80fd\u529b\uff08LRMs\uff09\u7684\u5174\u8d77\uff0cLLMs \u6b63\u4e0d\u65ad\u6269\u5c55\u5230\u591a\u6a21\u6001\u7406\u89e3\u3001\u751f\u6210\u4e0e\u590d\u6742\u63a8\u7406\u573a\u666f\u3002<\/p>\n\n\n\n<p>\u4f46\u6a21\u578b\u6027\u80fd\u6301\u7eed\u63d0\u5347\u7684\u80cc\u540e\uff0c\u662f\u6a21\u578b\u5c3a\u5bf8\u3001\u6570\u636e\u89c4\u6a21\u3001RL \u63a8\u7406\u957f\u5ea6\u7684\u5feb\u901f Scaling\uff0c\u662f\u7b97\u529b\u548c\u5b58\u50a8\u8d44\u6e90\u7684\u6025\u5267\u6d88\u8017\u3002\u5927\u6a21\u578b\u7684\u8bad\u7ec3\u4e0e\u63a8\u7406\u7684\u6210\u672c\u5c45\u9ad8\u4e0d\u4e0b\uff0c\u6210\u4e3a\u5236\u7ea6\u5176\u5e7f\u6cdb\u843d\u5730\u548c\u5e94\u7528\u7684\u73b0\u5b9e\u74f6\u9888\u3002<\/p>\n\n\n\n<p>\u672c\u6587\u4ece LLM \u67b6\u6784\u89d2\u5ea6\u51fa\u53d1\uff0c\u5e26\u4f60\u5256\u6790\u5927\u6a21\u578b\u7684\u6548\u7387\u79d8\u8bc0\u3002\u8fd9\u4e00\u5207\u7684\u6838\u5fc3\u5728\u4e8e Transformer \u67b6\u6784\u3002Transformer \u7684\u81ea\u6ce8\u610f\u529b\u673a\u5236\u867d\u5e26\u6765\u4e86\u8fdc\u8ddd\u79bb\u5efa\u6a21\u7684\u7a81\u7834\uff0c\u5374\u56e0 O(N2) \u7684\u590d\u6742\u5ea6\u5728\u957f\u5e8f\u5217\u4efb\u52a1\u4e2d\u6210\u672c\u9ad8\u6602\u3002<strong>\u800c\u5728 RAG\u3001\u667a\u80fd\u4f53\u3001\u957f\u94fe\u63a8\u7406\u3001\u591a\u6a21\u6001\u7b49\u65b0\u5174\u573a\u666f\u4e0b\uff0c\u957f\u5e8f\u5217\u9700\u6c42\u6108\u53d1\u7a81\u51fa\uff0c\u8fdb\u4e00\u6b65\u653e\u5927\u4e86\u6548\u7387\u4e0e\u6027\u80fd\u4e4b\u95f4\u7684\u77db\u76fe\u3002<\/strong>\u540c\u65f6 Transformer \u7684 FFN \u90e8\u5206\u91c7\u7528\u5bc6\u96c6\u7684 MLP \u5c42\uff0c\u540c\u6837\u9762\u4e34\u6a21\u578b\u89c4\u6a21\u653e\u5927\u540e\u7684\u8bad\u7ec3\u548c\u63a8\u7406\u6548\u7387\u95ee\u9898\u3002<\/p>\n\n\n\n<p>\u8fd1\u5e74\u6765\u9488\u5bf9 Transformer \u67b6\u6784\u6539\u8fdb\u7684\u521b\u65b0\u5de5\u4f5c\u5c42\u51fa\u4e0d\u7a77\uff0c\u5374\u4e00\u76f4\u7f3a\u4e4f\u4e00\u7bc7\u5168\u9762\u6df1\u5165\u7684\u7efc\u8ff0\u6587\u7ae0\u8fdb\u884c\u603b\u7ed3\u3002<\/p>\n\n\n\n<p>\u8fd1\u671f\uff0c<strong>\u4e0a\u6d77 AI Lab<\/strong>&nbsp;\u8054\u5408\u6e2f\u79d1\u5e7f\u3001\u6fb3\u95e8\u5927\u5b66\u3001\u4e2d\u79d1\u9662\u81ea\u52a8\u5316\u6240\u3001\u82cf\u5dde\u5927\u5b66\u3001\u745e\u5178 KTH\u3001\u5317\u5927\u3001\u6e2f\u4e2d\u6587\u7b49\u591a\u5bb6\u673a\u6784\uff0c\u603b\u7ed3 440 \u4f59\u7bc7\u76f8\u5173\u8bba\u6587\uff0c\u6df1\u5165\u63a2\u8ba8\u4e86\u5f53\u524d&nbsp;<strong>LLM <\/strong><strong>\u9ad8\u6548\u7ed3\u6784<\/strong>\u7684\u6700\u65b0\u8fdb\u5c55\uff0c\u5f62\u6210\u8fd9\u7bc7 82 \u9875\u7684\u7efc\u8ff0\u8bba\u6587\uff1a<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"335\" src=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-528-1024x335.jpg\" alt=\"\" class=\"wp-image-1346\" srcset=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-528-1024x335.jpg 1024w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-528-300x98.jpg 300w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-528-768x252.jpg 768w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-528.jpg 1108w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<ul class=\"wp-block-list\">\n<li>\u8bba\u6587\u6807\u9898\uff1aSpeed Always Wins: A Survey on Efficient Architectures for Large Language Models<\/li>\n\n\n\n<li>\u8bba\u6587\u5730\u5740\uff1ahttps:\/\/arxiv.org\/pdf\/2508.09834<\/li>\n\n\n\n<li>\u9879\u76ee\u4ed3\u5e93\uff1ahttps:\/\/github.com\/weigao266\/Awesome-Efficient-Arch<\/li>\n<\/ul>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"555\" src=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-530-1024x555.jpg\" alt=\"\" class=\"wp-image-1348\" srcset=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-530-1024x555.jpg 1024w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-530-300x163.jpg 300w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-530-768x417.jpg 768w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-530.jpg 1108w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<p><sup>&nbsp; &nbsp; &nbsp; \u56fe 2\uff1a\u5927\u8bed\u8a00\u6a21\u578b\u9ad8\u6548\u67b6\u6784\u6982\u89c8<\/sup><\/p>\n\n\n\n<p>\u8be5\u7efc\u8ff0\u5c06\u76ee\u524d LLM \u9ad8\u6548\u67b6\u6784\u603b\u7ed3\u5206\u7c7b\u4e3a\u4ee5\u4e0b 7 \u7c7b\uff1a<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>\u7ebf\u6027\u5e8f\u5217\u5efa\u6a21\uff1a\u964d\u4f4e\u6ce8\u610f\u529b\u8bad\u7ec3\u548c\u63a8\u7406\u590d\u6742\u5ea6\uff0c\u65e0\u9700 KV Cache \u5f00\u9500\u3002<\/strong><\/li>\n\n\n\n<li><strong>\u7a00\u758f\u5e8f\u5217\u5efa\u6a21\uff1a\u901a\u8fc7\u7a00\u758f\u5316\u6ce8\u610f\u529b\u77e9\u9635\uff0c\u964d\u4f4e\u8ba1\u7b97\u4e0e\u663e\u5b58\u9700\u6c42\u3002<\/strong><\/li>\n\n\n\n<li><strong>\u9ad8\u6548\u5168\u6ce8\u610f\u529b\uff1a\u5728\u4fdd\u6301\u5b8c\u6574\u6ce8\u610f\u529b\u7684\u524d\u63d0\u4e0b\u4f18\u5316\u5185\u5b58\u8bbf\u95ee\u4e0e KV \u5b58\u50a8\u3002<\/strong><\/li>\n\n\n\n<li><strong>\u7a00\u758f\u4e13\u5bb6\u6a21\u578b\uff1a\u901a\u8fc7\u6761\u4ef6\u6fc0\u6d3b\u90e8\u5206\u4e13\u5bb6\uff0c\u5927\u5e45\u63d0\u5347\u6a21\u578b\u5bb9\u91cf\u800c\u4e0d\u589e\u52a0\u7b49\u6bd4\u4f8b\u8ba1\u7b97\u6210\u672c\u3002<\/strong><\/li>\n\n\n\n<li><strong>\u6df7\u5408\u6a21\u578b\u67b6\u6784\uff1a\u7ed3\u5408\u7ebf\u6027\/\u7a00\u758f\u5e8f\u5217\u5efa\u6a21\u4e0e\u5168\u6ce8\u610f\u529b\uff0c\u517c\u987e\u6548\u7387\u4e0e\u6027\u80fd\u3002<\/strong><\/li>\n\n\n\n<li><strong>\u6269\u6563\u8bed\u8a00\u6a21\u578b\uff1a\u5229\u7528\u975e\u81ea\u56de\u5f52\u7684\u6269\u6563\u6a21\u578b\u8fdb\u884c\u8bed\u8a00\u751f\u6210\u3002<\/strong><\/li>\n\n\n\n<li><strong>\u5176\u4ed6\u6a21\u6001\u5e94\u7528\uff1a\u5c06\u8fd9\u4e9b\u9ad8\u6548\u67b6\u6784\u5e94\u7528\u4e8e\u89c6\u89c9\u3001\u8bed\u97f3\u3001\u591a\u6a21\u6001\u6a21\u578b\u3002<\/strong><\/li>\n<\/ul>\n\n\n\n<p>\u8fd9\u4e9b\u65b9\u5411\u7684\u63a2\u7d22\u4e0d\u4ec5\u5173\u4e4e LLM \u7684\u672a\u6765\u6548\u7387\uff0c\u4e5f\u5173\u4e4e\u5982\u4f55\u5728\u7b97\u529b\u53d7\u9650\u7684\u6761\u4ef6\u4e0b\uff0c\u6301\u7eed\u63a8\u52a8 AI \u8d70\u5411\u66f4\u5f3a\u7684\u667a\u80fd\u7684\u5173\u952e\u9009\u62e9\u3002\u7efc\u8ff0\u6d89\u53ca\u7684\u65b9\u6cd5\u7c7b\u522b\u548c\u4ee3\u8868\u6027\u8bba\u6587\u53ef\u89c1\u5982\u4e0b\u6811\u72b6\u56fe\uff1a<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"778\" height=\"1024\" src=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-527-778x1024.jpg\" alt=\"\" class=\"wp-image-1345\" srcset=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-527-778x1024.jpg 778w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-527-228x300.jpg 228w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-527-768x1011.jpg 768w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-527.jpg 1107w\" sizes=\"auto, (max-width: 778px) 100vw, 778px\" \/><\/figure>\n\n\n\n<p><sup>&nbsp; &nbsp; &nbsp; \u56fe 3\uff1a\u7efc\u8ff0\u5b8c\u6574\u7ec4\u7ec7\u67b6\u6784<\/sup><\/p>\n\n\n\n<p><strong>\u7ebf\u6027\u5e8f\u5217\u5efa\u6a21<\/strong><\/p>\n\n\n\n<p><strong>\u7ebf\u6027\u5e8f\u5217\u5efa\u6a21<\/strong>\u662f\u8fd1\u5e74\u6765\u7814\u7a76\u76f8\u5f53\u706b\u70ed\u7684\u4e00\u4e2a\u65b9\u5411\uff0c\u4ee3\u8868\u6027\u5de5\u4f5c\u50cf Mamba\u3001Lighting Attention\u3001RWKV\u3001GLA\u3001TTT \u7b49\u5728\u6a21\u578b\u67b6\u6784\u65b9\u5411\u90fd\u5f15\u8d77\u8fc7\u5e7f\u6cdb\u5173\u6ce8\u3002\u6211\u4eec\u5c06\u8fd9\u7c7b\u6280\u672f\u7ec6\u5206\u4e3a\u4ee5\u4e0b\u51e0\u4e2a\u7c7b\u522b\uff1a<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>\u7ebf\u6027\u6ce8\u610f\u529b<\/strong><\/li>\n\n\n\n<li><strong>\u7ebf\u6027 RNN<\/strong><\/li>\n\n\n\n<li><strong>\u72b6\u6001\u7a7a\u95f4\u6a21\u578b<\/strong><\/li>\n\n\n\n<li><strong>\u6d4b\u8bd5\u65f6\u63a8\u7406 RNN<\/strong><\/li>\n<\/ul>\n\n\n\n<p>\u5e76\u4e14\u6b63\u5982\u5728\u591a\u7bc7\u6587\u732e\u91cc\u5df2\u7ecf\u63d0\u51fa\u7684\uff0c\u8fd9\u4e9b\u7ebf\u6027\u5e8f\u5217\u5efa\u6a21\u65b9\u6cd5\u53ef\u4ee5\u6982\u62ec\u4e3a<strong>\u7edf\u4e00\u5efa\u6a21<\/strong>\u7684\u6570\u5b66\u5f62\u5f0f\uff0c\u5e76\u4e14\u80fd\u591f\u901a\u8fc7<strong>\u7ebf\u6027\u5316\u8fc7\u7a0b<\/strong>\u5c06\u9884\u8bad\u7ec3\u6a21\u578b\u6743\u91cd\u7684 Softmax Attention \u67b6\u6784\u8f6c\u4e3a Linear Sequence Modeling \u67b6\u6784\uff0c\u4ece\u800c\u83b7\u5f97\u6a21\u578b\u6548\u7387\u7684\u5927\u5e45\u63d0\u5347\uff0c\u5982\u4e0b\u56fe\u6240\u793a\u3002<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"503\" src=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-526-1024x503.jpg\" alt=\"\" class=\"wp-image-1344\" srcset=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-526-1024x503.jpg 1024w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-526-300x147.jpg 300w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-526-768x377.jpg 768w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-526.jpg 1108w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<p><sup>&nbsp; &nbsp; &nbsp; \u56fe 4\uff1a\u7ebf\u6027\u5e8f\u5217\u5efa\u6a21\u65b9\u6cd5<\/sup><\/p>\n\n\n\n<p>\u6211\u4eec\u5c06\u5df2\u6709\u7684\u7ebf\u6027\u5e8f\u5217\u5efa\u6a21\u65b9\u6cd5\u4ece\u8bb0\u5fc6\u89c6\u89d2\u548c\u4f18\u5316\u5668\u89c6\u89d2\u5206\u522b\u8fdb\u884c\u68b3\u7406\u548c\u5bf9\u6bd4\uff0c\u8be6\u7ec6\u5f62\u5f0f\u53ef\u89c1\u4e0b\u8868\uff1a<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"889\" src=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-529-1024x889.jpg\" alt=\"\" class=\"wp-image-1347\" srcset=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-529-1024x889.jpg 1024w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-529-300x260.jpg 300w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-529-768x667.jpg 768w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-529.jpg 1108w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<p><sup>&nbsp; &nbsp; &nbsp; \u8868 1\uff1a\u7ebf\u6027\u5e8f\u5217\u5efa\u6a21\u65b9\u6cd5\u7edf\u4e00\u5efa\u6a21\u7684 Memory \u89c6\u89d2\u548c Optimizer \u89c6\u89d2<\/sup><\/p>\n\n\n\n<p>\u5176\u4e2d<strong>\u7ebf\u6027\u5316\u6280\u672f<\/strong>\u53ef\u4ee5\u8fdb\u4e00\u6b65\u7ec6\u5206\u4e3a\u57fa\u4e8e\u5fae\u8c03\u7684\u7ebf\u6027\u5316\uff0c\u548c\u57fa\u4e8e\u84b8\u998f\u7684\u7ebf\u6027\u5316\uff0c\u5982\u4e0b\u56fe\u6240\u793a\uff1a<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"643\" src=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-531-1024x643.jpg\" alt=\"\" class=\"wp-image-1349\" srcset=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-531-1024x643.jpg 1024w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-531-300x188.jpg 300w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-531-768x482.jpg 768w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-531.jpg 1108w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<p><sup>&nbsp; &nbsp; &nbsp; \u56fe 5\uff1a\u7ebf\u6027\u5316\u65b9\u6cd5<\/sup><\/p>\n\n\n\n<p>\u7efc\u8ff0\u8fd8\u8fdb\u4e00\u6b65\u603b\u7ed3\u5f52\u7eb3\u4e86\u76ee\u524d\u5728\u7ebf\u6027\u5e8f\u5217\u5efa\u6a21\u9886\u57df\u5e38\u89c1\u7684<strong>\u786c\u4ef6\u9ad8\u6548\u5b9e\u73b0<\/strong>\u65b9\u6cd5\uff0c\u53ef\u4ee5\u5f52\u7eb3\u4e3a Blelloch Scan\u3001Chunk-wise Parallel \u548c Recurrent for Inferences\uff0c\u5982\u4e0b\u56fe\u6240\u793a\uff1a<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"468\" src=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-532-1024x468.jpg\" alt=\"\" class=\"wp-image-1350\" srcset=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-532-1024x468.jpg 1024w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-532-300x137.jpg 300w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-532-768x351.jpg 768w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-532.jpg 1108w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<p><sup>&nbsp; &nbsp; &nbsp; \u56fe 6\uff1a\u7ebf\u6027\u5e8f\u5217\u5efa\u6a21\u65b9\u6cd5\u7684\u786c\u4ef6\u9ad8\u6548\u5b9e\u73b0<\/sup><\/p>\n\n\n\n<p><strong>\u7a00\u758f\u5e8f\u5217\u5efa\u6a21<\/strong><\/p>\n\n\n\n<p><strong>\u7a00\u758f\u5e8f\u5217\u5efa\u6a21<\/strong>\u662f\u53e6\u4e00\u7c7b\u6709\u4ee3\u8868\u6027\u7684\u9ad8\u6548\u6ce8\u610f\u529b\u673a\u5236\uff0c\u901a\u8fc7\u5229\u7528 Attention Map \u5929\u7136\u5177\u6709\u7684\u7a00\u758f\u6027\u52a0\u901f\u6ce8\u610f\u529b\u7684\u8ba1\u7b97\uff0c\u8fd9\u7c7b\u65b9\u6cd5\u53ef\u4ee5\u8fdb\u4e00\u6b65\u7ec6\u5206\u4e3a\uff1a<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>\u9759\u6001\u7a00\u758f\u6ce8\u610f\u529b<\/strong><\/li>\n\n\n\n<li><strong>\u52a8\u6001\u7a00\u758f\u6ce8\u610f\u529b<\/strong><\/li>\n\n\n\n<li><strong>\u514d\u8bad\u7ec3\u7a00\u758f\u6ce8\u610f\u529b<\/strong><\/li>\n<\/ul>\n\n\n\n<p>\u4ee3\u8868\u6027\u7684\u7a00\u758f\u6ce8\u610f\u529b\u65b9\u6cd5\u5982 Global Attention\u3001Window Attention\u3001Dilated Attention \u7b49\uff0c\u53ca\u5176\u5de5\u4f5c\u539f\u7406\u5982\u4e0b\u56fe\u6240\u793a\uff1a<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"346\" src=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-533-1024x346.jpg\" alt=\"\" class=\"wp-image-1351\" srcset=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-533-1024x346.jpg 1024w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-533-300x101.jpg 300w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-533-768x259.jpg 768w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-533.jpg 1108w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<p><sup>&nbsp; &nbsp; &nbsp; \u56fe 7\uff1a\u7a00\u758f\u6ce8\u610f\u529b\u7684\u51e0\u79cd\u7ecf\u5178\u5f62\u5f0f<\/sup><\/p>\n\n\n\n<p><strong>\u9ad8\u6548\u5168\u6ce8\u610f\u529b<\/strong><\/p>\n\n\n\n<p>\u53e6\u4e00\u7c7b\u9ad8\u6548\u6ce8\u610f\u529b\u7b97\u6cd5\u53ef\u4ee5\u7edf\u4e00\u5f52\u7eb3\u4e3a<strong>\u9ad8\u6548\u5168\u6ce8\u610f\u529b<\/strong>\uff0c\u8fd9\u7c7b\u65b9\u6cd5\u53ef\u4ee5\u6839\u636e\u7b97\u6cd5\u601d\u8def\u8fdb\u4e00\u6b65\u7ec6\u5206\u4e3a\u5982\u4e0b\u51e0\u7c7b\uff1a<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>IO-Aware Attention<\/strong><\/li>\n\n\n\n<li><strong>Grouped Attention<\/strong><\/li>\n\n\n\n<li><strong>Mixture of Attention<\/strong><\/li>\n\n\n\n<li><strong>Quantized Attention<\/strong><\/li>\n<\/ul>\n\n\n\n<p>\u5176\u4e2d IO-Aware Attention \u6307\u4ee3\u76ee\u524d\u4f7f\u7528\u975e\u5e38\u5e7f\u6cdb\u7684 Flash Attention \u7cfb\u5217\u5de5\u4f5c\uff0cGrouped Attention \u5219\u5305\u542b\u5e7f\u4e3a\u4f7f\u7528\u7684 GQA\u3001MLA \u7b49\u5168\u6ce8\u610f\u529b\u53d8\u4f53\uff0c\u51e0\u79cd\u4ee3\u8868\u6027\u65b9\u6cd5\u5982\u4e0b\u56fe\u6240\u793a\u3002<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"787\" src=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-534-1024x787.jpg\" alt=\"\" class=\"wp-image-1352\" srcset=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-534-1024x787.jpg 1024w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-534-300x231.jpg 300w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-534-768x591.jpg 768w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-534.jpg 1108w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<p><sup>&nbsp; &nbsp; &nbsp; \u56fe 8\uff1aGrouped Attention \u7684\u51e0\u79cd\u4ee3\u8868\u6027\u65b9\u6cd5<\/sup><\/p>\n\n\n\n<p><strong>\u7a00\u758f\u6df7\u5408\u4e13\u5bb6<\/strong><\/p>\n\n\n\n<p><strong>\u7a00\u758f\u6df7\u5408\u4e13\u5bb6<\/strong>\u662f\u5bf9 Transformer \u67b6\u6784\u4e2d\u53e6\u4e00\u4e2a\u91cd\u8981\u6a21\u5757 FFN \u505a\u7684\u4e00\u7c7b\u91cd\u8981\u6539\u8fdb\uff0c\u5df2\u7ecf\u9010\u6e10\u6210\u4e3a\uff08\u8bed\u8a00\u548c\u591a\u6a21\u6001\uff09\u5927\u6a21\u578b\u67b6\u6784\u7684\u4e8b\u5b9e\u6807\u51c6\u3002\u7efc\u8ff0\u4e2d\u5c06\u76f8\u5173\u6587\u732e\u6309\u4ee5\u4e0b\u4e09\u4e2a\u65b9\u5411\u8fdb\u884c\u5206\u7c7b\uff1a<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>Routing Mechanisms<\/strong><\/li>\n\n\n\n<li><strong>Expert Architectures<\/strong><\/li>\n\n\n\n<li><strong>MoE Conversion<\/strong><\/li>\n<\/ul>\n\n\n\n<p>\u8def\u7531\u673a\u5236\u5305\u62ec Token-choice \u548c Expert-choice \u4e24\u7c7b\uff0c\u5176\u539f\u7406\u5982\u4e0b\u56fe\u6240\u793a\uff1a<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"357\" src=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-535-1024x357.jpg\" alt=\"\" class=\"wp-image-1353\" srcset=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-535-1024x357.jpg 1024w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-535-300x105.jpg 300w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-535-768x268.jpg 768w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-535.jpg 1108w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<p><sup>&nbsp; &nbsp; &nbsp; \u56fe 9\uff1aMoE \u8def\u7531\u673a\u5236<\/sup><\/p>\n\n\n\n<p>\u4e13\u5bb6\u7ed3\u6784\u7684\u521b\u65b0\u5de5\u4f5c\u5305\u62ec\uff1a\u5171\u4eab\u4e13\u5bb6\u3001\u7ec6\u7c92\u5ea6\u4e13\u5bb6\u3001\u96f6\u4e13\u5bb6\u3001\u6df1\u5ea6\u8def\u7531\u7b49\uff0c\u5176\u4f5c\u7528\u548c\u539f\u7406\u53ef\u89c1\u4e0b\u56fe\uff1a<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"565\" src=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-536-1024x565.jpg\" alt=\"\" class=\"wp-image-1354\" srcset=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-536-1024x565.jpg 1024w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-536-300x166.jpg 300w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-536-768x424.jpg 768w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-536.jpg 1107w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<p><sup>&nbsp; &nbsp; &nbsp; \u56fe 10\uff1aMoE \u4e13\u5bb6\u67b6\u6784<\/sup><\/p>\n\n\n\n<p>\u53e6\u5916\u4e00\u4e2a\u91cd\u8981\u7684\u65b9\u5411\u662f MoE \u8f6c\u6362\uff0c\u5df2\u6709\u7684\u5de5\u4f5c\u5305\u62ec\u901a\u8fc7 Split\u3001Copy\u3001Merge \u7b49\u624b\u6bb5\u5bf9\u4e13\u5bb6\u8fdb\u884c\u6784\u9020\uff0c\u5982\u4e0b\u56fe\u6240\u793a\uff1a<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"206\" src=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-537-1024x206.jpg\" alt=\"\" class=\"wp-image-1355\" srcset=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-537-1024x206.jpg 1024w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-537-300x60.jpg 300w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-537-768x155.jpg 768w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-537.jpg 1108w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<p><sup>&nbsp; &nbsp; &nbsp; \u56fe 11\uff1aMoE \u8f6c\u5316\u673a\u5236<\/sup><\/p>\n\n\n\n<p><strong>\u6df7\u5408\u67b6\u6784<\/strong><\/p>\n\n\n\n<p><strong>\u6df7\u5408\u67b6\u6784<\/strong>\u662f\u8fd1\u5e74\u6765\u51fa\u73b0\u7684\u4e00\u79cd\u5b9e\u7528\u7684\u65b0\u578b\u67b6\u6784\uff0c\u53ef\u4ee5\u5728\u7ebf\u6027\/\u7a00\u758f\u6ce8\u610f\u529b\u548c\u5168\u6ce8\u610f\u529b\u4e4b\u95f4\u53d6\u5f97\u5fae\u5999\u7684 Trade-off\uff0c\u4e5f\u5728\u6548\u7387\u548c\u6548\u679c\u95f4\u627e\u5230\u4e86\u6700\u4f73\u751c\u871c\u70b9\u3002\u5177\u4f53\u53ef\u7ec6\u5206\u4e3a\uff1a<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>\u5c42\u95f4\u6df7\u5408<\/strong><\/li>\n\n\n\n<li><strong>\u5c42\u5185\u6df7\u5408<\/strong><\/li>\n<\/ul>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"829\" src=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-538-1024x829.jpg\" alt=\"\" class=\"wp-image-1356\" srcset=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-538-1024x829.jpg 1024w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-538-300x243.jpg 300w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-538-768x622.jpg 768w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-538.jpg 1108w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<p><sup>&nbsp; &nbsp; &nbsp; \u56fe 12\uff1a\u6df7\u5408\u67b6\u6784\u5f62\u5f0f<\/sup><\/p>\n\n\n\n<p><strong>\u6269\u6563\u5927\u8bed\u8a00\u6a21\u578b<\/strong><\/p>\n\n\n\n<p><strong>\u6269\u6563\u5927\u8bed\u8a00\u6a21\u578b<\/strong>\u662f\u8fd1\u671f\u51fa\u73b0\u7684\u4e00\u4e2a\u70ed\u95e8\u65b9\u5411\uff0c\u521b\u65b0\u6027\u5730\u5c06\u6269\u6563\u6a21\u578b\u4ece\u89c6\u89c9\u751f\u6210\u4efb\u52a1\u8fc1\u79fb\u81f3\u8bed\u8a00\u4efb\u52a1\uff0c\u4ece\u800c\u5728\u8bed\u8a00\u751f\u6210\u901f\u5ea6\u4e0a\u53d6\u5f97\u5927\u5e45\u8fdb\u6b65\u3002\u76f8\u5173\u5de5\u4f5c\u53ef\u4ee5\u7ec6\u5206\u4e3a\uff1a<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>Non-Autoregressive Diffusion LLM<\/strong><\/li>\n\n\n\n<li><strong>Bridging Diffusion LLM and Autoregressive<\/strong><\/li>\n\n\n\n<li><strong>Extending Diffusion LLM to Multimodality<\/strong><\/li>\n<\/ul>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"466\" src=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-539-1024x466.jpg\" alt=\"\" class=\"wp-image-1357\" srcset=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-539-1024x466.jpg 1024w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-539-300x137.jpg 300w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-539-768x350.jpg 768w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-539.jpg 1107w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<p><sup>&nbsp; &nbsp; &nbsp; \u56fe 13\uff1a\u6269\u6563\u5927\u8bed\u8a00\u6a21\u578b\u673a\u5236<\/sup><\/p>\n\n\n\n<p><strong>\u5e94\u7528\u81f3\u5176\u4ed6\u6a21\u6001<\/strong><\/p>\n\n\n\n<p>\u6700\u540e\u4e00\u4e2a\u91cd\u8981\u7684\u90e8\u5206\u662f<strong>\u9ad8\u6548\u67b6\u6784\u5728\u5176\u4ed6\u6a21\u6001\u4e0a\u7684\u5e94\u7528\uff0c\u6db5\u76d6\u89c6\u89c9\u3001\u97f3\u9891\u548c\u591a\u6a21\u6001\u3002<\/strong>\u4ee5 Mamba \u4e3a\u4ee3\u8868\u7684\u7ebf\u6027\u6a21\u578b\u88ab\u5e7f\u6cdb\u5e94\u7528\u81f3\u591a\u79cd\u6a21\u6001\u4efb\u52a1\u4e0a\uff0c\u5e76\u53d6\u5f97\u4e86\u4f18\u79c0\u7684\u8868\u73b0\uff0c\u7efc\u8ff0\u5c06\u8fd9\u7c7b\u6a21\u578b\u603b\u7ed3\u68b3\u7406\u81f3\u5982\u4e0b\u8868\u683c\uff1a<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"652\" src=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-540-1024x652.jpg\" alt=\"\" class=\"wp-image-1358\" srcset=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-540-1024x652.jpg 1024w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-540-300x191.jpg 300w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-540-768x489.jpg 768w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-540.jpg 1108w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<p><strong>\u5bc4\u8bed<\/strong><\/p>\n\n\n\n<p>\u6700\u540e\u6b63\u5982\u5e1d\u56fd\u65f6\u4ee3 3 \u4e2d\u8fd9\u6761\u795e\u5947\u4ee3\u7801\u300cSpeed Always Wins\u300d\u80fd\u8ba9\u6e38\u620f\u4e16\u754c\u300cIncreases build, research, shipment, and gather rates by 100 times for all players\u300d\u4e00\u6837\uff0c\u6211\u4eec\u5e0c\u671b\u7efc\u8ff0\u4e2d\u7684 Efficient Architectures \u53ef\u4ee5\u771f\u88ab\u7528\u6765 100x \u52a0\u901f AI \u4e16\u754c\u7684\u6784\u5efa\uff0c\u66f4\u5feb\u66f4\u4fbf\u5b9c\u5730\u8bad\u7ec3\u51fa\u66f4\u5f3a\u66f4\u5b9e\u7528\u7684\u5927\u6a21\u578b\u3002<\/p>\n","protected":false},"excerpt":{"rendered":"<p>\u00a0\u56fe 1\uff1a\u5e38\u89c1\u957f\u5e8f\u5217\u573a\u666f \u8fd1\u5e74\u6765\uff0c\u5927\u8bed\u8a00\u6a21\u578b\uff08LLMs\uff09\u5c55\u73b0\u51fa\u5f3a\u5927\u7684\u8bed\u8a00\u7406\u89e3\u4e0e\u751f\u6210\u80fd\u529b\uff0c\u63a8\u52a8\u4e86\u6587\u672c\u751f\u6210\u3001\u4ee3\u7801\u751f [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":1343,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[10],"tags":[],"class_list":["post-1342","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-10"],"_links":{"self":[{"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/posts\/1342","targetHints":{"allow":["GET"]}}],"collection":[{"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/comments?post=1342"}],"version-history":[{"count":1,"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/posts\/1342\/revisions"}],"predecessor-version":[{"id":1359,"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/posts\/1342\/revisions\/1359"}],"wp:featuredmedia":[{"embeddable":true,"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/media\/1343"}],"wp:attachment":[{"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/media?parent=1342"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/categories?post=1342"},{"taxonomy":"post_tag","embeddable":true,"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/tags?post=1342"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}