问世两周年看ChatGPT发展史:从基础神经元到多模态智能体_腾讯新闻

############### \u003cdiv class="rich_media_content autoTypeSetting24psection"\u003e\u003csection data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="margin-bottom: 0px"\u003e\u003c/p\u003e\u003csection style="margin-bottom: unset" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="margin-bottom: 10px; margin-top: 10px" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="background-color: rgb(255, 213, 34); border-radius: 50%; height: 50px; line-height: 0; margin-bottom: unset; overflow: hidden; width: 50px" data-exeditor-arbitrary-box="wrap"\u003e\u003cp\u003e\u003cspan style="background-color: transparent"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003csection style="background-color: rgb(254, 254, 254); box-shadow: rgb(216, 175, 127) 0px 0px 3px; margin: -40px 10px; padding: 15px; transform: rotateZ(0deg)" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="background-color: transparent; text-indent: 2em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="background-color: transparent"\u003e本文从基本概念出发,介绍和解释ChatGPT用到的一系列关键技术,如机器学习、神经网络、大模型、预训练+微调范式、Scaling Law……希望帮助读者更为深入地了解和使用以ChatGPT为代表的相关工具,助力读者成为人工智能时代的弄潮儿。\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003csection style="background-color: rgb(255, 213, 34); border-radius: 50%; height: 50px; line-height: 0; margin-bottom: unset; margin-left: auto; overflow: hidden; width: 50px" data-exeditor-arbitrary-box="wrap"\u003e\u003cp\u003e\u003cspan style="background-color: transparent"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003c/section\u003e\u003cp style="margin-bottom: 0px"\u003e\u003cspan style="letter-spacing: 2px"\u003e\u003cspan style="font-size: 0px"\u003e\u003cspan style="color: rgba(255, 255, 255, 0)"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003cp style="margin-bottom: 0px"\u003e\u003cspan style="letter-spacing: 2px"\u003e\u003cspan style="font-size: 0px"\u003e\u003cspan style="color: rgba(255, 255, 255, 0)"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003cp style="color: rgba(255, 255, 255, 0); font-size: 0px; letter-spacing: 2px; margin-bottom: 0px" data-exeditor-arbitrary-box="image-box"\u003e\u003c!--IMG_0--\u003e\u003c/p\u003e\u003c/section\u003e\u003csection style="margin-bottom: 0px" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp\u003e\u003cspan style="font-size: 16px"\u003e2022年11月30日,一家名不见经传的公司(OpenAI)悄悄上线了一个产品ChatGPT。彼时,谁也没有想到这款产品会在短短几个月内风靡全球;\u003cstrong\u003e\u003cspan style="color: rgb(0, 82, 255)"\u003e而2023年3月14日GPT-4的发布更是激起了一场属于生成式人工智能(artificial intelligence generated content, AIGC)的科技革命。\u003c/span\u003e\u003c/strong\u003e对于普通人来说,面对这个正在给生产和生活带来巨大改变的人工智能产品,不禁会产生无数的疑问:\u003c/span\u003e\u003c!--MID_AD_0--\u003e\u003c!--EOP_0--\u003e\u003c/p\u003e\u003c!--MID_ARTICLE_AD_0--\u003e\u003c!--PARAGRAPH_0--\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="margin-bottom: 0px" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="color: rgb(99, 125, 223); font-size: 14px; letter-spacing: 2px; line-height: 1.75em; margin-bottom: unset; padding: 10px; transform: rotateZ(0deg)" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 24px; text-indent: 2em"\u003e\u003cspan style="letter-spacing: 2px"\u003e\u003cspan style="font-size: 15px"\u003e\u003cspan style="color: rgb(136, 136, 136)"\u003e●ChatGPT为什么引起如此大的重视?\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003cp style="line-height: 24px; text-indent: 2em"\u003e\u003cspan style="letter-spacing: 2px"\u003e\u003cspan style="font-size: 15px"\u003e\u003cspan style="color: rgb(136, 136, 136)"\u003e●它的原理是什么?\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003cp style="line-height: 24px; text-indent: 2em"\u003e\u003cspan style="letter-spacing: 2px"\u003e\u003cspan style="font-size: 15px"\u003e\u003cspan style="color: rgb(136, 136, 136)"\u003e●它真的具备人类的智慧吗?\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003cp style="line-height: 24px; text-indent: 2em"\u003e\u003cspan style="letter-spacing: 2px"\u003e\u003cspan style="font-size: 15px"\u003e\u003cspan style="color: rgb(136, 136, 136)"\u003e●它将给人类社会带来哪些变化?\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003cp style="line-height: 24px; text-indent: 2em"\u003e\u003cspan style="letter-spacing: 2px"\u003e\u003cspan style="font-size: 15px"\u003e\u003cspan style="color: rgb(121, 123, 170)"\u003e……\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="margin-bottom: 0px" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="margin-bottom: 20px; margin-top: 10px; text-align: center" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="display: inline-block; margin-bottom: unset" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="background-color: rgb(253, 201, 56); border-radius: 50%; height: 34px; line-height: 0; margin-bottom: unset; margin-left: auto; overflow: hidden; width: 34px" data-exeditor-arbitrary-box="wrap"\u003e\u003cp\u003e\u003cspan style="background-color: transparent"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003csection style="background-color: rgb(254, 254, 254); border-color: rgb(126, 148, 230); border-style: solid; border-width: 1px 1px 4px; margin-bottom: unset; margin-left: 8px; margin-right: 12px; margin-top: -22px" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="background-color: transparent; color: rgb(99, 125, 223); letter-spacing: 2px; margin-bottom: unset; padding: 5px 20px" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="background-color: transparent"\u003e\u003cspan style="letter-spacing: 0px"\u003e\u003cspan style="color: rgb(99, 125, 223)"\u003e\u003cspan style="background-color: transparent"\u003eChatGPT原理概览:\u003cbr/\u003e文字接龙游戏\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="background-color: rgb(253, 201, 56); height: 6px; line-height: 0; margin-bottom: unset; margin-top: -16px; overflow: hidden; width: 18px" data-exeditor-arbitrary-box="wrap"\u003e\u003cp\u003e\u003cspan style="background-color: transparent"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cstrong\u003e\u003cspan style="color: rgb(0, 82, 255)"\u003eChatGPT最令人印象深刻的能力是它能够通过对话的方式回答用户的问题,那么ChatGPT回答问题的原理是什么呢?\u003c/span\u003e\u003c/strong\u003e传统的问答系统本质上是基于数据库和搜索引擎,即通过搜索引擎在网络与数据库中搜索相关信息,然后把结果直接返回给用户。比如我们使用百度搜索“机器学习的原理是什么”,百度会跳转出各式各样的网站。这些网站是由各个企业早就开发好的,百度仅仅是根据相关度做了一个排序。\u003c/span\u003e\u003c/p\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e不同于传统问答系统中答案来源于现成的网络或者数据库,ChatGPT的回答是随着提问的进行自动生成的。\u003c/span\u003e\u003cstrong\u003e\u003cspan style="color: rgb(0, 82, 255)"\u003e这一点有点像文字接龙游戏,ChatGPT会基于前面的话不断地生成下一个合适的词汇,直到觉得不必继续生成为止。\u003c/span\u003e\u003c/strong\u003e\u003c/span\u003e\u003c/p\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e比如我们问ChatGPT:“苹果是一种水果吗”,ChatGPT会基于这句话进行文字接龙,大概流程如下:\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e(1)考虑下一个可能的词汇及其对应的概率,如右表(为了方便理解只写了3个可能的形式)所示。\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003cp style="line-height: 24px; text-indent: 2em"\u003e\u003cbr/\u003e\u003c/p\u003e\u003cp style="text-align: center" data-exeditor-arbitrary-box="image-box"\u003e\u003c!--IMG_1--\u003e\u003c/p\u003e\u003cp style="line-height: 24px; text-indent: 2em"\u003e\u003cbr/\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e(2)基于上述概率分布,ChatGPT会选择概率最大的答案,即“是的”(因为其概率0.8明显大于其他选项)。\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e(3)此时这句话的内容变成 “苹果是一种水果么?是的”,ChatGPT会看下一个可能的词和对应概率是什么。\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e不断重复这个步骤,直到得到一个完整的回答。\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e从上面例子可以看出:\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e(1)不同于传统问答基于数据库或搜索引擎,ChatGPT的答案是在用户输入问题以后,随着问题自动生成的。\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e(2)这种生成本质上是在做文字接龙,简单来说是不断在所有可能词汇中选择概率最大的词汇来生成。\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cstrong\u003e\u003cspan style="color: rgb(0, 82, 255)"\u003e有些聪明的读者会有一个疑问,ChatGPT是怎么知道该选择什么词汇,又是如何给出各个可能词汇的概率呢?\u003c/span\u003e\u003c/strong\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e这正是\u003c!--SECURE_LINK_BEGIN_0--\u003e机器学习\u003c!--SECURE_LINK_END_0--\u003e技术的神奇之处。\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="margin-bottom: 0px" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="margin-bottom: 20px; margin-top: 10px; text-align: center" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="display: inline-block; margin-bottom: unset" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="background-color: rgb(253, 201, 56); border-radius: 50%; height: 34px; line-height: 0; margin-bottom: unset; margin-left: auto; overflow: hidden; width: 34px" data-exeditor-arbitrary-box="wrap"\u003e\u003cp\u003e\u003cspan style="background-color: transparent"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003csection style="background-color: rgb(254, 254, 254); border-color: rgb(126, 148, 230); border-style: solid; border-width: 1px 1px 4px; margin-bottom: unset; margin-left: 8px; margin-right: 12px; margin-top: -22px" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="background-color: transparent; color: rgb(99, 125, 223); letter-spacing: 2px; margin-bottom: unset; padding: 5px 20px" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="background-color: transparent"\u003e\u003cspan style="letter-spacing: 0px"\u003e\u003cspan style="color: rgb(99, 125, 223)"\u003e\u003cspan style="background-color: transparent"\u003e机器学习的核心:\u003cbr/\u003e模仿人类进行学习\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="background-color: rgb(253, 201, 56); height: 6px; line-height: 0; margin-bottom: unset; margin-top: -16px; overflow: hidden; width: 18px" data-exeditor-arbitrary-box="wrap"\u003e\u003cp\u003e\u003cspan style="background-color: transparent"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cstrong\u003e\u003cspan style="color: rgb(0, 82, 255)"\u003eChatGPT是机器学习的一个非常典型的应用,那么什么是机器学习呢?\u003c/span\u003e\u003c/strong\u003e\u003c/span\u003e\u003c/p\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cstrong\u003e\u003cspan style="color: rgb(0, 82, 255)"\u003e机器学习整体思想是借鉴人类学习的过程。\u003c/span\u003e\u003c/strong\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e人类观察、归纳客观世界的实际情况,并从中学到相关的规律,当面对某一未知情况的时候,会使用已经学到的规律来解决未知的问题。同理,我们希望计算机能够自动地从海量数据中发现某种“规律”,并将这种规律应用于一些新的问题。\u003c/span\u003e\u003cstrong\u003e\u003cspan style="color: rgb(0, 82, 255)"\u003e这种规律在机器学习领域就被称为“模型”,学习的过程被称为对模型进行训练。\u003c/span\u003e\u003c/strong\u003e\u003c/span\u003e\u003c/p\u003e\u003cp style="line-height: 24px; text-indent: 2em"\u003e\u003cbr/\u003e\u003c/p\u003e\u003cp style="text-align: center" data-exeditor-arbitrary-box="image-box"\u003e\u003c!--IMG_2--\u003e\u003c/p\u003e\u003cp style="text-align: center"\u003e\u003cspan style="font-size: 14px"\u003e\u003cspan style="color: rgb(136, 136, 136)"\u003e机器学习和模型训练卡通示意图\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003cp style="line-height: 24px; text-indent: 2em"\u003e\u003cbr/\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cstrong\u003e\u003cspan style="color: rgb(0, 82, 255)"\u003e关于模型训练,实际上所有机器学习模型背后都有一个假设:学习的规律是能够通过数学表示的。\u003c/span\u003e\u003c/strong\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e机器学习的核心就是想办法找到一个数学函数,让这个函数尽可能接近真实世界的数学表达式。然而很多时候人类并不知道真实的数学表示是什么形式,也无法通过传统数学推导的方式获得;人类唯一拥有的是一堆来源于真实情境的数据。机器学习的方法就是使用这些数据(训练数据)去训练我们的模型,让模型自动找到一个较好的近似结果。比如人脸识别的应用,就是想找到一个函数,这个函数的输入是人脸照片,输出是判定这张照片对应哪个人。\u003c/span\u003e\u003cstrong\u003e\u003cspan style="color: rgb(0, 82, 255)"\u003e然而人类不知道人脸识别函数是什么形式,于是就拿来一大堆人脸的照片并且标记好每个脸对应的人,交给模型去训练,让模型自动找到一个较好的人脸识别函数。这就是机器学习在做的事情。\u003c/span\u003e\u003c/strong\u003e\u003c/span\u003e\u003c!--MID_AD_1--\u003e\u003c!--EOP_1--\u003e\u003c/p\u003e\u003c!--MID_ARTICLE_AD_1--\u003e\u003c!--PARAGRAPH_1--\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="margin-bottom: 0px" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="margin-bottom: 20px; margin-top: 10px; text-align: center" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="display: inline-block; margin-bottom: unset" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="background-color: rgb(253, 201, 56); border-radius: 50%; height: 34px; line-height: 0; margin-bottom: unset; margin-left: auto; overflow: hidden; width: 34px" data-exeditor-arbitrary-box="wrap"\u003e\u003cp\u003e\u003cspan style="background-color: transparent"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003csection style="background-color: rgb(254, 254, 254); border-color: rgb(126, 148, 230); border-style: solid; border-width: 1px 1px 4px; margin-bottom: unset; margin-left: 8px; margin-right: 12px; margin-top: -22px" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="background-color: transparent; color: rgb(99, 125, 223); letter-spacing: 2px; margin-bottom: unset; padding: 5px 20px" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="background-color: transparent"\u003e\u003cspan style="letter-spacing: 0px"\u003e\u003cspan style="color: rgb(99, 125, 223)"\u003e\u003cspan style="background-color: transparent"\u003e神经网络与神经元:\u003cbr/\u003e可扩展的数学表达能力\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="background-color: rgb(253, 201, 56); height: 6px; line-height: 0; margin-bottom: unset; margin-top: -16px; overflow: hidden; width: 18px" data-exeditor-arbitrary-box="wrap"\u003e\u003cp\u003e\u003cspan style="background-color: transparent"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e理解了机器学习是什么,另一个概念是机器学习模型的数学表达能力。机器学习模型本质上是想要尽可能接近真实世界对应的那个函数。然而正如我们不能指望仅仅通过几条直线就画出精美绝伦的美术作品,\u003c/span\u003e\u003cstrong\u003e\u003cspan style="color: rgb(0, 82, 255)"\u003e如果机器学习模型本身比较简单,比如高中学到的线性函数\u003c/span\u003e\u003c/strong\u003e\u003c/span\u003e\u003c/p\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-align: center; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003eY=kx+b\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e那么它无论如何也不可能学习出一个复杂的函数。因此机器学习模型的一个重要考虑点就是模型的数学表达能力,当面对一个复杂问题的时候,我们希望模型数学表达能力尽可能强,这样模型才有可能学好。\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cstrong\u003e\u003cspan style="color: rgb(0, 82, 255)"\u003e过去几十年科学家发明了非常多不同的机器学习模型,而其中最具影响力的是一种叫作“神经网络”的模型。\u003c/span\u003e\u003c/strong\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e神经网络模型最初基于生物学的一个现象:人类神经元的基础架构非常简单,只能做一些基础的信号处理工作,但最终通过大脑能够完成复杂的思考。\u003c/span\u003e\u003cstrong\u003e\u003cspan style="color: rgb(0, 82, 255)"\u003e受此启发,科学家们开始思考是否可以构建一些简单的“神经元”,并通过神经元的连接形成网络,从而产生处理复杂信息的能力。\u003c/span\u003e\u003c/strong\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e基于此,神经网络的基础单元是一个神经元的模型,其只能进行简单的计算。假设输入数据有2个维度(x\u003csub\u003e1\u003c/sub\u003e, x\u003csub\u003e2\u003c/sub\u003e),那么这个神经网络可以写成\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-align: center; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003ey=σ(w\u003csub\u003e1\u003c/sub\u003ex\u003csub\u003e1\u003c/sub\u003e+w\u003csub\u003e2\u003c/sub\u003ex\u003csub\u003e2\u003c/sub\u003e+b)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003cp style="line-height: 24px; text-indent: 2em"\u003e\u003cspan style="color: rgb(121, 123, 170)"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/p\u003e\u003cp data-exeditor-arbitrary-box="image-box"\u003e\u003c!--IMG_3--\u003e\u003c/p\u003e\u003csection style="text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp\u003e\u003cspan style="font-size: 14px"\u003e\u003cspan style="color: rgb(136, 136, 136)"\u003e从神经元到神经网络  (a)神经元架构(生物);(b)基础神经元架构(人工智能);(c)简单多层感知机。\u003cbr/\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003cp\u003e\u003cspan style="color: rgb(121, 123, 170)"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cstrong\u003e\u003cspan style="color: rgb(0, 82, 255)"\u003e上述神经元的数学表达能力非常弱,只是一个简单的线性函数和一个激活函数的组合;但是我们可以很轻松地把模型变得强大起来,方案就是增加更多的“隐藏节点”。\u003c/span\u003e\u003c/strong\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e在这个时候虽然每个节点依然进行非常简单的计算,但组合起来其数学表达能力就会变得很强。感兴趣的读者可以尝试类比上述公式写出下图中简单多层感知机对应的公式,将会得到一个非常复杂的公式。\u003c/span\u003e\u003cstrong\u003e\u003cspan style="color: rgb(0, 82, 255)"\u003e这个模型也是日后深度学习的基础模型,即多层感知机[1]。\u003c/span\u003e\u003c/strong\u003e\u003c/span\u003e\u003c/p\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e多层感知机的原理非常简单,但是透过它可以很好地了解神经网络的原理:虽然单个神经元非常简单,但是通过大量节点的组合就可以让模型具备非常强大的数学表达能力。而之后整个深度学习的技术路线,某种程度上就是沿着开发并训练更大更深的网络的路线前进的。\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="margin-bottom: 0px" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="margin-bottom: 20px; margin-top: 10px; text-align: center" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="display: inline-block; margin-bottom: unset" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="background-color: rgb(253, 201, 56); border-radius: 50%; height: 34px; line-height: 0; margin-bottom: unset; margin-left: auto; overflow: hidden; width: 34px" data-exeditor-arbitrary-box="wrap"\u003e\u003cp\u003e\u003cspan style="background-color: transparent"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003csection style="background-color: rgb(254, 254, 254); border-color: rgb(126, 148, 230); border-style: solid; border-width: 1px 1px 4px; margin-bottom: unset; margin-left: 8px; margin-right: 12px; margin-top: -22px" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="background-color: transparent; color: rgb(99, 125, 223); letter-spacing: 2px; margin-bottom: unset; padding: 5px 20px" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="background-color: transparent"\u003e\u003cspan style="letter-spacing: 0px"\u003e\u003cspan style="color: rgb(99, 125, 223)"\u003e\u003cspan style="background-color: transparent"\u003e深度学习新范式:\u003cbr/\u003e预训练+微调范式与Scaling Law\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="background-color: rgb(253, 201, 56); height: 6px; line-height: 0; margin-bottom: unset; margin-top: -16px; overflow: hidden; width: 18px" data-exeditor-arbitrary-box="wrap"\u003e\u003cp\u003e\u003cspan style="background-color: transparent"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cstrong\u003e\u003cspan style="color: rgb(0, 82, 255)"\u003e深度学习领域从2012年开始蓬勃发展,更大更深且效果更好的模型不断出现。然而随着模型越来越复杂,从头训练模型的成本越来越高。\u003c/span\u003e\u003c/strong\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e于是有人提出,能否不从头训练,而是在别人训练好的模型基础上训练,从而用更低的成本达到更好的效果呢?\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cstrong\u003e\u003cspan style="color: rgb(0, 82, 255)"\u003e例如,科学家对一个图像分类模型进行拆分,希望研究深度学习模型里的那么多层都学到了什么东西[2]。\u003c/span\u003e\u003c/strong\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e结果发现,越接近输入层,模型学到的是越基础的信息,比如边、角、纹理等;越接近输出层,\u003c/span\u003e\u003cstrong\u003e\u003cspan style="color: rgb(0, 82, 255)"\u003e模型学到的是越接近高级组合的信息,比如公鸡的形状、船的形状等。\u003c/span\u003e\u003c/strong\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e不仅仅在图像领域如此,在自然语言、语音等很多领域也存在这个特征。\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003cp style="line-height: 24px; text-indent: 2em"\u003e\u003cbr/\u003e\u003c/p\u003e\u003cp style="text-align: center" data-exeditor-arbitrary-box="image-box"\u003e\u003c!--IMG_4--\u003e\u003c/p\u003e\u003csection style="text-align: justify; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp\u003e\u003cspan style="font-size: 14px"\u003e\u003cspan style="color: rgb(136, 136, 136)"\u003e深度神经网络中不同层的输出  接近输入层(左侧)一般是基础信息,接近输出层(右侧)一般是某个具体的物体等信息[2]。\u003c/span\u003e\u003c/span\u003e\u003cbr/\u003e\u003c/p\u003e\u003c/section\u003e\u003cp style="line-height: 24px; text-indent: 2em"\u003e\u003cbr/\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cstrong\u003e\u003cspan style="color: rgb(0, 82, 255)"\u003e基础信息往往是领域通用的信息,比如图像领域的边、角、纹理等,在各类图像识别中都会用到;\u003c/span\u003e\u003c/strong\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e而高级组合信息往往是领域专用信息,比如猫的形状只有在动物识别任务中才有用,在人脸识别的任务就没用。\u003c/span\u003e\u003cstrong\u003e\u003cspan style="color: rgb(0, 82, 255)"\u003e因此一个自然而然的逻辑是,通过领域常见数据训练出一个通用的模型,主要是学好领域通用信息;\u003c/span\u003e\u003c/strong\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e在面对某个具体场景时,只需要使用该场景数据做个小规模训练(微调)就可以了。\u003c/span\u003e\u003cstrong\u003e\u003cspan style="color: rgb(0, 82, 255)"\u003e这就是著名的预训练+微调的范式。\u003c/span\u003e\u003c/strong\u003e\u003c/span\u003e\u003c/p\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cstrong\u003e\u003cspan style="color: rgb(0, 82, 255)"\u003e预训练+微调这一范式的出现与普及对领域产生了两个重大影响。\u003c/span\u003e\u003c/strong\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e一方面,在已有模型基础上微调大大降低了成本;另一方面,一个好的预训练模型的重要性也更加凸显,因此各大公司、科研机构更加愿意花大量成本来训练更加昂贵的基础模型。那么大模型的效果到底与什么因素有关呢?\u003c!--SECURE_LINK_BEGIN_1--\u003eOpenAI\u003c!--SECURE_LINK_END_1--\u003e在2020年提出了著名的Scaling Law,即当模型规模变大以后,模型的效果主要受到模型参数规模、训练数据规模和使用算力规模影响[3]。\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cstrong\u003e\u003cspan style="color: rgb(0, 82, 255)"\u003eScaling Law积极的一面是为提升模型效果指明了方向\u003c/span\u003e\u003c/strong\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e,只要把模型和数据规模做得更大就可以,这也是为什么近年来大模型的规模在以指数级增长,以及基础算力资源图形处理器(graphics processing unit,  GPU)总是供不应求;但Scaling Law也揭示了一个让很多科学家绝望的事实:即模型的每一步提升都需要人类用极为夸张的算力和数据成本来“交换”。大模型的成本门槛变得非常之高,从头训练大模型成了学界的奢望,以OpenAI、谷歌、Meta、百度、智谱AI等企业为代表的业界开始发挥引领作用。\u003c/span\u003e\u003c/span\u003e\u003c!--MID_AD_2--\u003e\u003c!--EOP_2--\u003e\u003c/p\u003e\u003c!--MID_ARTICLE_AD_2--\u003e\u003c!--PARAGRAPH_2--\u003e\u003cp style="line-height: 24px; text-indent: 2em"\u003e\u003cbr/\u003e\u003c/p\u003e\u003cp style="text-align: center" data-exeditor-arbitrary-box="image-box"\u003e\u003c!--IMG_5--\u003e\u003c/p\u003e\u003cp style="line-height: 24px; text-indent: 2em"\u003e\u003cbr/\u003e\u003c/p\u003e\u003cp style="line-height: 24px; text-indent: 2em"\u003e\u003cbr/\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="margin-bottom: 0px" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="margin-bottom: 20px; margin-top: 10px; text-align: center" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="display: inline-block; margin-bottom: unset" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="background-color: rgb(253, 201, 56); border-radius: 50%; height: 34px; line-height: 0; margin-bottom: unset; margin-left: auto; overflow: hidden; width: 34px" data-exeditor-arbitrary-box="wrap"\u003e\u003cp\u003e\u003cspan style="background-color: transparent"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003csection style="background-color: rgb(254, 254, 254); border-color: rgb(126, 148, 230); border-style: solid; border-width: 1px 1px 4px; margin-bottom: unset; margin-left: 8px; margin-right: 12px; margin-top: -22px" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="background-color: transparent; color: rgb(99, 125, 223); letter-spacing: 2px; margin-bottom: unset; padding: 5px 20px" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="background-color: transparent"\u003e\u003cspan style="letter-spacing: 0px"\u003e\u003cspan style="color: rgb(99, 125, 223)"\u003e\u003cspan style="background-color: transparent"\u003eGPT的野心:\u003cbr/\u003e上下文学习与提示词工程\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="background-color: rgb(253, 201, 56); height: 6px; line-height: 0; margin-bottom: unset; margin-top: -16px; overflow: hidden; width: 18px" data-exeditor-arbitrary-box="wrap"\u003e\u003cp\u003e\u003cspan style="background-color: transparent"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cstrong\u003e\u003cspan style="color: rgb(0, 82, 255)"\u003e除了希望通过训练规模巨大的模型来提升效果以外,GPT模型在发展过程中还有一个非常雄大的野心:上下文学习(in-context learning)。\u003c/span\u003e\u003c/strong\u003e\u003c/span\u003e\u003c/p\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cstrong\u003e\u003cspan style="color: rgb(0, 82, 255)"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/strong\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e正如前文所述,在过去如果想要模型“学”到什么内容,需要用一大堆数据来训练我们的模型;哪怕是前文讲到的预训练+微调的范式,依然需要在已训练好的模型基础上,用一个小批量数据做训练(即微调)。因此在过去,“训练”一直是机器学习中最核心的概念。但OpenAI提出,训练本身既有成本又有门槛,希望模型面对新任务的时候不用额外训练,只需要在对话窗口里给模型一些例子,模型就自动学会了。这种模式就叫作上下文学习。\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e举一个中英文翻译的例子。过去做中英文翻译,需要使用海量的中英文数据集训练一个机器学习模型;而在上下文学习中,想要完成同样的任务,只需要给模型一些例子,比如告诉模型下面的话:\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="color: rgb(99, 125, 223); font-size: 14px; letter-spacing: 2px; line-height: 24px; margin-bottom: unset; padding: 10px; text-indent: 2em; transform: rotateZ(0deg)" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="color: rgb(99, 125, 223); font-size: 14px; letter-spacing: 2px; line-height: 24px; margin-bottom: unset; padding: 10px; text-indent: 2em; transform: rotateZ(0deg)" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="color: rgb(99, 125, 223); font-size: 14px; letter-spacing: 2px; line-height: 24px; margin-bottom: unset; padding: 10px; text-indent: 2em; transform: rotateZ(0deg)" data-exeditor-arbitrary-box="wrap"\u003e\u003cp\u003e\u003cspan style="letter-spacing: 2px"\u003e\u003cspan style="font-size: 15px"\u003e\u003cspan style="color: rgb(136, 136, 136)"\u003e下面是一些中文翻译成英文的例子:\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003csection style="color: rgb(99, 125, 223); font-size: 14px; letter-spacing: 2px; line-height: 24px; margin-bottom: unset; padding: 10px; text-indent: 2em; transform: rotateZ(0deg)" data-exeditor-arbitrary-box="wrap"\u003e\u003cp\u003e\u003cspan style="letter-spacing: 2px"\u003e\u003cspan style="font-size: 15px"\u003e\u003cspan style="color: rgb(136, 136, 136)"\u003e我爱中国 → I love China\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003csection style="color: rgb(99, 125, 223); font-size: 14px; letter-spacing: 2px; line-height: 24px; margin-bottom: unset; padding: 10px; text-indent: 2em; transform: rotateZ(0deg)" data-exeditor-arbitrary-box="wrap"\u003e\u003cp\u003e\u003cspan style="letter-spacing: 2px"\u003e\u003cspan style="font-size: 15px"\u003e\u003cspan style="color: rgb(136, 136, 136)"\u003e我喜欢写代码 → I love coding\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003csection style="color: rgb(99, 125, 223); font-size: 14px; letter-spacing: 2px; line-height: 24px; margin-bottom: unset; padding: 10px; text-indent: 2em; transform: rotateZ(0deg)" data-exeditor-arbitrary-box="wrap"\u003e\u003cp\u003e\u003cspan style="letter-spacing: 2px"\u003e\u003cspan style="font-size: 15px"\u003e\u003cspan style="color: rgb(136, 136, 136)"\u003e\u003c!--SECURE_LINK_BEGIN_2--\u003e人工智能\u003c!--SECURE_LINK_END_2--\u003e很重要 → AI is important\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003csection style="color: rgb(99, 125, 223); font-size: 14px; letter-spacing: 2px; line-height: 24px; margin-bottom: unset; padding: 10px; text-indent: 2em; transform: rotateZ(0deg)" data-exeditor-arbitrary-box="wrap"\u003e\u003cp\u003e\u003cspan style="letter-spacing: 2px"\u003e\u003cspan style="font-size: 15px"\u003e\u003cspan style="color: rgb(136, 136, 136)"\u003e现在我有一句中文,请翻译成英文。这句话是:“我今天想吃苹果”。\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e这时候原本“傻傻的”模型就突然具备了翻译的能力,能够自动翻译了。\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e有过ChatGPT使用经历的读者会发现,这个输入就是提示词(prompt)。在ChatGPT使用已相当普及的今天,很多人意识不到这件事有多神奇。\u003c/span\u003e\u003cstrong\u003e\u003cspan style="color: rgb(0, 82, 255)"\u003e这就如同找一个没学过英语的孩子,给他看几个中英文翻译的句子,这个孩子就可以流畅地进行中英文翻译了。\u003c/span\u003e\u003c/strong\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e要知道这个模型可从来没有专门在中英文翻译的数据集上训练过,也就是说模型本身并没有中英文翻译的能力,但它竟然通过对话里的一些例子就突然脱胎换骨“顿悟”了中英文翻译,这真的非常神奇!\u003c/span\u003e\u003c/span\u003e\u003c!--MID_AD_3--\u003e\u003c!--EOP_3--\u003e\u003c/p\u003e\u003c!--MID_ARTICLE_AD_3--\u003e\u003c!--PARAGRAPH_3--\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cstrong\u003e\u003cspan style="color: rgb(0, 82, 255)"\u003e上下文学习的相关机制到今天依然是学界讨论的热点,而恰恰因为GPT模型具有上下文学习的能力,一个好的提示词非常重要。\u003c/span\u003e\u003c/strong\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e提示词工程逐步成为一个热门的领域,甚至出现了一种新的职业叫作“提示词工程师”(prompt engineer),就是通过写出更好的提示词让ChatGPT发挥更大的作用。\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cstrong\u003e\u003cspan style="color: rgb(0, 82, 255)"\u003eChatGPT原理总结如下:\u003c/span\u003e\u003c/strong\u003e\u003c/span\u003e\u003c/p\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e(1) ChatGPT本质是在做文字接龙的游戏,在游戏中它会根据候选词汇的概率来挑选下一个词。\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e(2) ChatGPT背后是一个非常庞大的神经网络,比如GPT-3有1700亿个参数(训练成本在100万美元以上)。\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e(3)基于庞大的神经网络,面对一句话时,模型可以准确给出候选词汇的概率,从而完成文字接龙的操作。\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e(4)这种有巨大规模进行语言处理的模型,也叫作大语言模型(large language model)。\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="line-height: 1.6em; margin: 0px; text-indent: 0em" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="line-height: 1.6em; margin-bottom: 0px; margin-left: 0px; margin-right: 0px; margin-top: 0px; text-indent: 0em"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(0, 0, 0)"\u003e(5)以GPT为代表的大语言模型具备上下文学习的能力,因此一个好的提示词至关重要。\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="margin-bottom: 0px" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="color: rgb(99, 125, 223); font-size: 14px; letter-spacing: 2px; line-height: 1.75em; margin-bottom: unset; padding: 10px; transform: rotateZ(0deg)" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="background-color: rgb(255, 255, 255); letter-spacing: 0.544px; line-height: 2em; margin: 16px 8px 0px; outline: 0px" data-exeditor-arbitrary-box="wrap"\u003e\u003cp\u003e\u003cspan style="letter-spacing: 0.544px"\u003e\u003cspan style="font-size: 14px"\u003e\u003cstrong\u003e\u003cspan style="color: rgb(123, 127, 131)"\u003e\u003cspan style="background-color: transparent"\u003e参考文献:\u003c/span\u003e\u003c/span\u003e\u003c/strong\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003csection style="background-color: rgb(255, 255, 255); color: rgb(34, 34, 34); font-size: 16px; letter-spacing: 0.544px; margin-bottom: 0px; outline: 0px" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="background-color: transparent; letter-spacing: 0.544px; outline: 0px" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="background-color: transparent; letter-spacing: 0.544px; outline: 0px" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="background-color: transparent; line-height: 1.75em; margin-bottom: 8px; margin-left: 8px; margin-right: 8px; outline: 0px; text-align: left; width: 677px" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="background-color: transparent; display: inline-block; outline: 0px; vertical-align: top; width: 677px" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="background-color: transparent; outline: 0px; overflow-x: hidden" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="background-color: transparent; margin-top: 20px; outline: 0px" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="background-color: transparent; border-color: rgb(252, 251, 251); border-style: solid; border-width: 2px; display: inline-block; outline: 0px; padding: 0.1em 0.3em" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="background-color: transparent; letter-spacing: 0.578px; margin-bottom: 0px" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="background-color: transparent; color: rgb(99, 125, 223); font-size: 14px; letter-spacing: 2px; line-height: 1.75em; margin-bottom: unset; padding: 10px; transform: rotateZ(0deg)" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="background-color: transparent; line-height: 24px; text-align: left"\u003e\u003cspan style="letter-spacing: 2px"\u003e\u003cspan style="font-size: 14px"\u003e\u003cspan style="color: rgb(136, 136, 136)"\u003e\u003cspan style="background-color: transparent"\u003e[1] Rumelhart D E, Hinton E G, Williams R J. Learning representations by back-propagating errors. Nature, 1986, 323(6088): 533–536.\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="background-color: transparent; letter-spacing: 0.578px; margin-bottom: 0px" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="background-color: transparent; color: rgb(99, 125, 223); font-size: 14px; letter-spacing: 2px; line-height: 1.75em; margin-bottom: unset; padding: 10px; transform: rotateZ(0deg)" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="background-color: transparent; line-height: 24px; text-align: left"\u003e\u003cspan style="letter-spacing: 2px"\u003e\u003cspan style="font-size: 14px"\u003e\u003cspan style="color: rgb(136, 136, 136)"\u003e\u003cspan style="background-color: transparent"\u003e[2] Wei D L, Zhou B L, Torralba A,et al. mNeuron: A matlab plugin to visualize neurons from deep models. (2015)[2024-07-05]. https://donglaiw.github.io/proj/mneuron/index.html.\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="background-color: transparent; letter-spacing: 0.578px; margin-bottom: 0px" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="background-color: transparent; color: rgb(99, 125, 223); font-size: 14px; letter-spacing: 2px; line-height: 1.75em; margin-bottom: unset; padding: 10px; transform: rotateZ(0deg)" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="background-color: transparent; line-height: 24px; text-align: left"\u003e\u003cspan style="letter-spacing: 2px"\u003e\u003cspan style="font-size: 14px"\u003e\u003cspan style="color: rgb(136, 136, 136)"\u003e\u003cspan style="background-color: transparent"\u003e[3] Kaplan J, McCandlish S, Henighan T, et al. Scaling laws for neural language models. arXiv preprint arXiv: 2001.08361, 2020.\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="background-color: transparent; letter-spacing: 0.578px; margin-bottom: 0px" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="background-color: transparent; color: rgb(99, 125, 223); font-size: 14px; letter-spacing: 2px; line-height: 1.75em; margin-bottom: unset; padding: 10px; transform: rotateZ(0deg)" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="background-color: transparent; line-height: 24px; text-align: left"\u003e\u003cspan style="letter-spacing: 2px"\u003e\u003cspan style="font-size: 14px"\u003e\u003cspan style="color: rgb(136, 136, 136)"\u003e\u003cspan style="background-color: transparent"\u003e[4] Zhao W X, Zhou K, Li J, et al. A survey of large language models. arXiv preprint arXiv: 2303.18223, 2023.\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="background-color: transparent; letter-spacing: 0.578px; margin-bottom: 0px" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="background-color: transparent; color: rgb(99, 125, 223); font-size: 14px; letter-spacing: 2px; line-height: 1.75em; margin-bottom: unset; padding: 10px; transform: rotateZ(0deg)" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="background-color: transparent; line-height: 24px; text-align: left"\u003e\u003cspan style="letter-spacing: 2px"\u003e\u003cspan style="font-size: 14px"\u003e\u003cspan style="color: rgb(136, 136, 136)"\u003e\u003cspan style="background-color: transparent"\u003e[5] Boiko D A, MacKnight R, Kline B, et al. Autonomous chemical research with large language models. Nature, 2023, 624(7992): 570-578.\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="background-color: transparent; letter-spacing: 0.578px; margin-bottom: 0px" data-exeditor-arbitrary-box="wrap"\u003e\u003csection style="background-color: transparent; color: rgb(99, 125, 223); font-size: 14px; letter-spacing: 2px; line-height: 1.75em; margin-bottom: unset; padding: 10px; transform: rotateZ(0deg)" data-exeditor-arbitrary-box="wrap"\u003e\u003cp style="background-color: transparent; line-height: 24px; text-align: left"\u003e\u003cspan style="letter-spacing: 2px"\u003e\u003cspan style="font-size: 14px"\u003e\u003cspan style="color: rgb(136, 136, 136)"\u003e\u003cspan style="background-color: transparent"\u003e关键词:ChatGPT   机器学习   神经网络   大模型   多模态智能体\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003cp style="background-color: transparent; line-height: 2em; margin-bottom: 32px; margin-left: 8px; margin-right: 8px"\u003e\u003cspan style="letter-spacing: 0.544px"\u003e\u003cspan style="font-size: 16px"\u003e\u003cspan style="color: rgb(34, 34, 34)"\u003e\u003cspan style="background-color: transparent"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003c/section\u003e\u003c/section\u003e\u003c/section\u003e\u003c/section\u003e\u003c/section\u003e\u003c!--MID_AD_4--\u003e\u003c!--EOP_4--\u003e\u003c/section\u003e\u003c!--MID_ARTICLE_AD_4--\u003e\u003c!--PARAGRAPH_4--\u003e\u003cp style="line-height: 24px; text-align: left"\u003e\u003cspan style="letter-spacing: 2px"\u003e\u003cspan style="font-size: 14px"\u003e\u003cspan style="color: rgb(121, 123, 170)"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003c/section\u003e\u003csection style="letter-spacing: 0.578px; margin-bottom: 0px" data-exeditor-arbitrary-box="wrap"\u003e\u003chr style="border-bottom-width: 0px; border-color: rgba(0, 0, 0, 0.1); border-left-width: 0px; border-right-width: 0px; border-style: solid; transform: scale(1, 0.5); transform-origin: 0px 0px"/\u003e\u003cp\u003e\u003cspan style="letter-spacing: 0.544px"\u003e\u003cspan style="font-size: 14px"\u003e\u003cstrong\u003e\u003cspan style="color: rgb(136, 136, 136)"\u003e来源:科学杂志1915\u003c/span\u003e\u003c/strong\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003cp\u003e\u003cspan style="letter-spacing: 0.544px"\u003e\u003cspan style="font-size: 14px"\u003e\u003cstrong\u003e\u003cspan style="color: rgb(136, 136, 136)"\u003e原标题:ChatGPT发展史:从基础神经元到多模态智能体\u003c/span\u003e\u003c/strong\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan style="letter-spacing: 0.578px"\u003e\u003cbr/\u003e\u003c/span\u003e\u003c/p\u003e\u003cp style="background-color: rgb(255, 255, 255); text-align: right"\u003e\u003cspan style="letter-spacing: 0.544px"\u003e\u003cspan style="font-size: 14px"\u003e\u003cstrong\u003e\u003cspan style="color: rgb(136, 136, 136)"\u003e\u003cspan style="background-color: rgb(255, 255, 255)"\u003e编辑:未\u003c/span\u003e\u003c/span\u003e\u003c/strong\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003c/section\u003e\u003cstyle\u003e.data_color_scheme_dark{--weui-BTN-ACTIVE-MASK: rgba(255, 255, 255, .1)}.data_color_scheme_dark{--weui-BTN-DEFAULT-ACTIVE-BG: rgba(255, 255, 255, .126)}.data_color_scheme_dark{--weui-DIALOG-LINE-COLOR: rgba(255, 255, 255, .1)}.data_color_scheme_dark{--weui-BG-COLOR-ACTIVE: #373737}.data_color_scheme_dark{--weui-BG-6: rgba(255, 255, 255, .1);--weui-ACTIVE-MASK: rgba(255, 255, 255, .1)}.data_color_scheme_dark{--weui-BG-0: #111;--weui-BG-1: #1e1e1e;--weui-BG-5: #2c2c2c;--weui-RED: #fa5151;--weui-ORANGERED: #ff6146;--weui-ORANGE: #c87d2f;--weui-YELLOW: #cc9c00;--weui-GREEN: #74a800;--weui-LIGHTGREEN: #3eb575;--weui-BRAND: #07c160;--weui-BLUE: #10aeff;--weui-INDIGO: #1196ff;--weui-PURPLE: #8183ff;--weui-LINK: #7d90a9;--weui-TEXTGREEN: #259c5c;--weui-REDORANGE: #ff6146;--weui-BG-0: #111111;--weui-BG-1: #1E1E1E;--weui-BG-2: #191919;--weui-BG-3: #202020;--weui-BG-4: #404040;--weui-BG-5: #2C2C2C;--weui-BLUE-100: #10AEFF;--weui-BLUE-120: #0C8BCC;--weui-BLUE-170: #04344D;--weui-BLUE-80: #3FBEFF;--weui-BLUE-90: #28B6FF;--weui-BLUE-BG-100: #48A6E2;--weui-BLUE-BG-110: #4095CB;--weui-BLUE-BG-130: #32749E;--weui-BLUE-BG-90: #5AAFE4;--weui-BRAND-100: #07C160;--weui-BRAND-120: #059A4C;--weui-BRAND-170: #023A1C;--weui-BRAND-80: #38CD7F;--weui-BRAND-90: #20C770;--weui-BRAND-BG-100: #2AAE67;--weui-BRAND-BG-110: #259C5C;--weui-BRAND-BG-130: #1D7A48;--weui-BRAND-BG-90: #3EB575;--weui-FG-0: rgba(255, 255, 255, .8);--weui-FG-0_5: rgba(255, 255, 255, .6);--weui-FG-1: rgba(255, 255, 255, .5);--weui-FG-2: rgba(255, 255, 255, .3);--weui-FG-3: rgba(255, 255, 255, .1);--weui-FG-4: rgba(255, 255, 255, .15);--weui-GLYPH-0: rgba(255, 255, 255, .8);--weui-GLYPH-1: rgba(255, 255, 255, .5);--weui-GLYPH-2: rgba(255, 255, 255, .3);--weui-GLYPH-WHITE-0: rgba(255, 255, 255, .8);--weui-GLYPH-WHITE-1: rgba(255, 255, 255, .5);--weui-GLYPH-WHITE-2: rgba(255, 255, 255, .3);--weui-GLYPH-WHITE-3: #FFFFFF;--weui-GREEN-100: #74A800;--weui-GREEN-120: #5C8600;--weui-GREEN-170: #233200;--weui-GREEN-80: #8FB933;--weui-GREEN-90: #82B01A;--weui-GREEN-BG-100: #789833;--weui-GREEN-BG-110: #6B882D;--weui-GREEN-BG-130: #65802B;--weui-GREEN-BG-90: #85A247;--weui-INDIGO-100: #1196FF;--weui-INDIGO-120: #0D78CC;--weui-INDIGO-170: #052D4D;--weui-INDIGO-80: #40ABFF;--weui-INDIGO-90: #28A0FF;--weui-INDIGO-BG-100: #0D78CC;--weui-INDIGO-BG-110: #0B6BB7;--weui-INDIGO-BG-130: #09548F;--weui-INDIGO-BG-90: #2585D1;--weui-LIGHTGREEN-100: #3EB575;--weui-LIGHTGREEN-120: #31905D;--weui-LIGHTGREEN-170: #123522;--weui-LIGHTGREEN-80: #64C390;--weui-LIGHTGREEN-90: #51BC83;--weui-LIGHTGREEN-BG-100: #31905D;--weui-LIGHTGREEN-BG-110: #2C8153;--weui-LIGHTGREEN-BG-130: #226541;--weui-LIGHTGREEN-BG-90: #31905D;--weui-LINK-100: #7D90A9;--weui-LINK-120: #647387;--weui-LINK-170: #252A32;--weui-LINK-80: #97A6BA;--weui-LINK-90: #899AB1;--weui-LINKFINDER-100: #DEE9FF;--weui-MATERIAL-ATTACHMENTCOLUMN: rgba(32, 32, 32, .93);--weui-MATERIAL-NAVIGATIONBAR: rgba(18, 18, 18, .9);--weui-MATERIAL-REGULAR: rgba(37, 37, 37, .6);--weui-MATERIAL-THICK: rgba(34, 34, 34, .9);--weui-MATERIAL-THIN: rgba(95, 95, 95, .4);--weui-MATERIAL-TOOLBAR: rgba(35, 35, 35, .93);--weui-ORANGE-100: #C87D2F;--weui-ORANGE-120: #A06425;--weui-ORANGE-170: #3B250E;--weui-ORANGE-80: #D39758;--weui-ORANGE-90: #CD8943;--weui-ORANGE-BG-100: #BB6000;--weui-ORANGE-BG-110: #A85600;--weui-ORANGE-BG-130: #824300;--weui-ORANGE-BG-90: #C1701A;--weui-ORANGERED-100: #FF6146;--weui-OVERLAY: rgba(0, 0, 0, .8);--weui-OVERLAY-WHITE: rgba(242, 242, 242, .8);--weui-PURPLE-100: #8183FF;--weui-PURPLE-120: #6768CC;--weui-PURPLE-170: #26274C;--weui-PURPLE-80: #9A9BFF;--weui-PURPLE-90: #8D8FFF;--weui-PURPLE-BG-100: #6768CC;--weui-PURPLE-BG-110: #5C5DB7;--weui-PURPLE-BG-130: #48498F;--weui-PURPLE-BG-90: #7677D1;--weui-RED-100: #FA5151;--weui-RED-120: #C84040;--weui-RED-170: #4B1818;--weui-RED-80: #FB7373;--weui-RED-90: #FA6262;--weui-RED-BG-100: #CF5148;--weui-RED-BG-110: #BA4940;--weui-RED-BG-130: #913832;--weui-RED-BG-90: #D3625A;--weui-SECONDARY-BG: rgba(255, 255, 255, .1);--weui-SEPARATOR-0: rgba(255, 255, 255, .05);--weui-SEPARATOR-1: rgba(255, 255, 255, .15);--weui-STATELAYER-HOVERED: rgba(0, 0, 0, .02);--weui-STATELAYER-PRESSED: rgba(255, 255, 255, .1);--weui-STATELAYER-PRESSEDSTRENGTHENED: rgba(255, 255, 255, .2);--weui-YELLOW-100: #CC9C00;--weui-YELLOW-120: #A37C00;--weui-YELLOW-170: #3D2F00;--weui-YELLOW-80: #D6AF33;--weui-YELLOW-90: #D1A519;--weui-YELLOW-BG-100: #BF9100;--weui-YELLOW-BG-110: #AB8200;--weui-YELLOW-BG-130: #866500;--weui-YELLOW-BG-90: #C59C1A;--weui-FG-HALF: rgba(255, 255, 255, .6);--weui-RED: #FA5151;--weui-ORANGERED: #FF6146;--weui-ORANGE: #C87D2F;--weui-YELLOW: #CC9C00;--weui-GREEN: #74A800;--weui-LIGHTGREEN: #3EB575;--weui-TEXTGREEN: #259C5C;--weui-BRAND: #07C160;--weui-BLUE: #10AEFF;--weui-INDIGO: #1196FF;--weui-PURPLE: #8183FF;--weui-LINK: #7D90A9;--weui-REDORANGE: #FF6146;--weui-TAG-TEXT-BLACK: rgba(255, 255, 255, .5);--weui-TAG-BACKGROUND-BLACK: rgba(255, 255, 255, .05);--weui-WHITE: rgba(255, 255, 255, .8);--weui-FG: #fff;--weui-BG: #000;--weui-FG-5: rgba(255, 255, 255, .1);--weui-TAG-BACKGROUND-ORANGE: rgba(250, 157, 59, .1);--weui-TAG-BACKGROUND-GREEN: rgba(6, 174, 86, .1);--weui-TAG-TEXT-RED: rgba(250, 81, 81, .6);--weui-TAG-BACKGROUND-RED: rgba(250, 81, 81, .1);--weui-TAG-BACKGROUND-BLUE: rgba(16, 174, 255, .1);--weui-TAG-TEXT-ORANGE: rgba(250, 157, 59, .6);--weui-TAG-TEXT-GREEN: rgba(6, 174, 86, .6);--weui-TAG-TEXT-BLUE: rgba(16, 174, 255, .6)}.data_color_scheme_dark{--weui-BG-0: #111111;--weui-BG-1: #1E1E1E;--weui-BG-2: #191919;--weui-BG-3: #202020;--weui-BG-4: #404040;--weui-BG-5: #2C2C2C;--weui-BLUE-100: #10AEFF;--weui-BLUE-120: #0C8BCC;--weui-BLUE-170: #04344D;--weui-BLUE-80: #3FBEFF;--weui-BLUE-90: #28B6FF;--weui-BLUE-BG-100: #48A6E2;--weui-BLUE-BG-110: #4095CB;--weui-BLUE-BG-130: #32749E;--weui-BLUE-BG-90: #5AAFE4;--weui-BRAND-100: #07C160;--weui-BRAND-120: #059A4C;--weui-BRAND-170: #023A1C;--weui-BRAND-80: #38CD7F;--weui-BRAND-90: #20C770;--weui-BRAND-BG-100: #2AAE67;--weui-BRAND-BG-110: #259C5C;--weui-BRAND-BG-130: #1D7A48;--weui-BRAND-BG-90: #3EB575;--weui-FG-0: rgba(255, 255, 255, .8);--weui-FG-0_5: rgba(255, 255, 255, .6);--weui-FG-1: rgba(255, 255, 255, .5);--weui-FG-2: rgba(255, 255, 255, .3);--weui-FG-3: rgba(255, 255, 255, .1);--weui-FG-4: rgba(255, 255, 255, .15);--weui-GLYPH-0: rgba(255, 255, 255, .8);--weui-GLYPH-1: rgba(255, 255, 255, .5);--weui-GLYPH-2: rgba(255, 255, 255, .3);--weui-GLYPH-WHITE-0: rgba(255, 255, 255, .8);--weui-GLYPH-WHITE-1: rgba(255, 255, 255, .5);--weui-GLYPH-WHITE-2: rgba(255, 255, 255, .3);--weui-GLYPH-WHITE-3: #FFFFFF;--weui-GREEN-100: #74A800;--weui-GREEN-120: #5C8600;--weui-GREEN-170: #233200;--weui-GREEN-80: #8FB933;--weui-GREEN-90: #82B01A;--weui-GREEN-BG-100: #789833;--weui-GREEN-BG-110: #6B882D;--weui-GREEN-BG-130: #65802B;--weui-GREEN-BG-90: #85A247;--weui-INDIGO-100: #1196FF;--weui-INDIGO-120: #0D78CC;--weui-INDIGO-170: #052D4D;--weui-INDIGO-80: #40ABFF;--weui-INDIGO-90: #28A0FF;--weui-INDIGO-BG-100: #0D78CC;--weui-INDIGO-BG-110: #0B6BB7;--weui-INDIGO-BG-130: #09548F;--weui-INDIGO-BG-90: #2585D1;--weui-LIGHTGREEN-100: #3EB575;--weui-LIGHTGREEN-120: #31905D;--weui-LIGHTGREEN-170: #123522;--weui-LIGHTGREEN-80: #64C390;--weui-LIGHTGREEN-90: #51BC83;--weui-LIGHTGREEN-BG-100: #31905D;--weui-LIGHTGREEN-BG-110: #2C8153;--weui-LIGHTGREEN-BG-130: #226541;--weui-LIGHTGREEN-BG-90: #31905D;--weui-LINK-100: #7D90A9;--weui-LINK-120: #647387;--weui-LINK-170: #252A32;--weui-LINK-80: #97A6BA;--weui-LINK-90: #899AB1;--weui-LINKFINDER-100: #DEE9FF;--weui-MATERIAL-ATTACHMENTCOLUMN: rgba(32, 32, 32, .93);--weui-MATERIAL-NAVIGATIONBAR: rgba(18, 18, 18, .9);--weui-MATERIAL-REGULAR: rgba(37, 37, 37, .6);--weui-MATERIAL-THICK: rgba(34, 34, 34, .9);--weui-MATERIAL-THIN: rgba(95, 95, 95, .4);--weui-MATERIAL-TOOLBAR: rgba(35, 35, 35, .93);--weui-ORANGE-100: #C87D2F;--weui-ORANGE-120: #A06425;--weui-ORANGE-170: #3B250E;--weui-ORANGE-80: #D39758;--weui-ORANGE-90: #CD8943;--weui-ORANGE-BG-100: #BB6000;--weui-ORANGE-BG-110: #A85600;--weui-ORANGE-BG-130: #824300;--weui-ORANGE-BG-90: #C1701A;--weui-ORANGERED-100: #FF6146;--weui-OVERLAY: rgba(0, 0, 0, .8);--weui-OVERLAY-WHITE: rgba(242, 242, 242, .8);--weui-PURPLE-100: #8183FF;--weui-PURPLE-120: #6768CC;--weui-PURPLE-170: #26274C;--weui-PURPLE-80: #9A9BFF;--weui-PURPLE-90: #8D8FFF;--weui-PURPLE-BG-100: #6768CC;--weui-PURPLE-BG-110: #5C5DB7;--weui-PURPLE-BG-130: #48498F;--weui-PURPLE-BG-90: #7677D1;--weui-RED-100: #FA5151;--weui-RED-120: #C84040;--weui-RED-170: #4B1818;--weui-RED-80: #FB7373;--weui-RED-90: #FA6262;--weui-RED-BG-100: #CF5148;--weui-RED-BG-110: #BA4940;--weui-RED-BG-130: #913832;--weui-RED-BG-90: #D3625A;--weui-SECONDARY-BG: rgba(255, 255, 255, .1);--weui-SEPARATOR-0: rgba(255, 255, 255, .05);--weui-SEPARATOR-1: rgba(255, 255, 255, .15);--weui-STATELAYER-HOVERED: rgba(0, 0, 0, .02);--weui-STATELAYER-PRESSED: rgba(255, 255, 255, .1);--weui-STATELAYER-PRESSEDSTRENGTHENED: rgba(255, 255, 255, .2);--weui-YELLOW-100: #CC9C00;--weui-YELLOW-120: #A37C00;--weui-YELLOW-170: #3D2F00;--weui-YELLOW-80: #D6AF33;--weui-YELLOW-90: #D1A519;--weui-YELLOW-BG-100: #BF9100;--weui-YELLOW-BG-110: #AB8200;--weui-YELLOW-BG-130: #866500;--weui-YELLOW-BG-90: #C59C1A;--weui-FG-HALF: rgba(255, 255, 255, .6);--weui-RED: #FA5151;--weui-ORANGERED: #FF6146;--weui-ORANGE: #C87D2F;--weui-YELLOW: #CC9C00;--weui-GREEN: #74A800;--weui-LIGHTGREEN: #3EB575;--weui-TEXTGREEN: #259C5C;--weui-BRAND: #07C160;--weui-BLUE: #10AEFF;--weui-INDIGO: #1196FF;--weui-PURPLE: #8183FF;--weui-LINK: #7D90A9;--weui-REDORANGE: #FF6146;--weui-TAG-TEXT-BLACK: rgba(255, 255, 255, .5);--weui-TAG-BACKGROUND-BLACK: rgba(255, 255, 255, .05);--weui-WHITE: rgba(255, 255, 255, .8);--weui-FG: #fff;--weui-BG: #000;--weui-FG-5: rgba(255, 255, 255, .1);--weui-TAG-BACKGROUND-ORANGE: rgba(250, 157, 59, .1);--weui-TAG-BACKGROUND-GREEN: rgba(6, 174, 86, .1);--weui-TAG-TEXT-RED: rgba(250, 81, 81, .6);--weui-TAG-BACKGROUND-RED: rgba(250, 81, 81, .1);--weui-TAG-BACKGROUND-BLUE: rgba(16, 174, 255, .1);--weui-TAG-TEXT-ORANGE: rgba(250, 157, 59, .6);--weui-TAG-TEXT-GREEN: rgba(6, 174, 86, .6);--weui-TAG-TEXT-BLUE: rgba(16, 174, 255, .6)}.data_color_scheme_dark{--weui-elpsColor: rgba(255, 255, 255, .8)}.data_color_scheme_dark{--weui-mask-elpsColor: rgba(255, 255, 255, .8);--weui-mask-gradient: linear-gradient(to right, rgba(25, 25, 25, 0), #191919 40%)}.data_color_scheme_dark{--weui-BG-0: #111111;--weui-BG-1: #1E1E1E;--weui-BG-2: #191919;--weui-BG-3: #202020;--weui-BG-4: #404040;--weui-BG-5: #2C2C2C;--weui-BLUE-100: #10AEFF;--weui-BLUE-120: #0C8BCC;--weui-BLUE-170: #04344D;--weui-BLUE-80: #3FBEFF;--weui-BLUE-90: #28B6FF;--weui-BLUE-BG-100: #48A6E2;--weui-BLUE-BG-110: #4095CB;--weui-BLUE-BG-130: #32749E;--weui-BLUE-BG-90: #5AAFE4;--weui-BRAND-100: #07C160;--weui-BRAND-120: #059A4C;--weui-BRAND-170: #023A1C;--weui-BRAND-80: #38CD7F;--weui-BRAND-90: #20C770;--weui-BRAND-BG-100: #2AAE67;--weui-BRAND-BG-110: #259C5C;--weui-BRAND-BG-130: #1D7A48;--weui-BRAND-BG-90: #3EB575;--weui-FG-0: rgba(255, 255, 255, .8);--weui-FG-0_5: rgba(255, 255, 255, .6);--weui-FG-1: rgba(255, 255, 255, .5);--weui-FG-2: rgba(255, 255, 255, .3);--weui-FG-3: rgba(255, 255, 255, .1);--weui-FG-4: rgba(255, 255, 255, .15);--weui-GLYPH-0: rgba(255, 255, 255, .8);--weui-GLYPH-1: rgba(255, 255, 255, .5);--weui-GLYPH-2: rgba(255, 255, 255, .3);--weui-GLYPH-WHITE-0: rgba(255, 255, 255, .8);--weui-GLYPH-WHITE-1: rgba(255, 255, 255, .5);--weui-GLYPH-WHITE-2: rgba(255, 255, 255, .3);--weui-GLYPH-WHITE-3: #FFFFFF;--weui-GREEN-100: #74A800;--weui-GREEN-120: #5C8600;--weui-GREEN-170: #233200;--weui-GREEN-80: #8FB933;--weui-GREEN-90: #82B01A;--weui-GREEN-BG-100: #789833;--weui-GREEN-BG-110: #6B882D;--weui-GREEN-BG-130: #65802B;--weui-GREEN-BG-90: #85A247;--weui-INDIGO-100: #1196FF;--weui-INDIGO-120: #0D78CC;--weui-INDIGO-170: #052D4D;--weui-INDIGO-80: #40ABFF;--weui-INDIGO-90: #28A0FF;--weui-INDIGO-BG-100: #0D78CC;--weui-INDIGO-BG-110: #0B6BB7;--weui-INDIGO-BG-130: #09548F;--weui-INDIGO-BG-90: #2585D1;--weui-LIGHTGREEN-100: #3EB575;--weui-LIGHTGREEN-120: #31905D;--weui-LIGHTGREEN-170: #123522;--weui-LIGHTGREEN-80: #64C390;--weui-LIGHTGREEN-90: #51BC83;--weui-LIGHTGREEN-BG-100: #31905D;--weui-LIGHTGREEN-BG-110: #2C8153;--weui-LIGHTGREEN-BG-130: #226541;--weui-LIGHTGREEN-BG-90: #31905D;--weui-LINK-100: #7D90A9;--weui-LINK-120: #647387;--weui-LINK-170: #252A32;--weui-LINK-80: #97A6BA;--weui-LINK-90: #899AB1;--weui-LINKFINDER-100: #DEE9FF;--weui-MATERIAL-ATTACHMENTCOLUMN: rgba(32, 32, 32, .93);--weui-MATERIAL-NAVIGATIONBAR: rgba(18, 18, 18, .9);--weui-MATERIAL-REGULAR: rgba(37, 37, 37, .6);--weui-MATERIAL-THICK: rgba(34, 34, 34, .9);--weui-MATERIAL-THIN: rgba(95, 95, 95, .4);--weui-MATERIAL-TOOLBAR: rgba(35, 35, 35, .93);--weui-ORANGE-100: #C87D2F;--weui-ORANGE-120: #A06425;--weui-ORANGE-170: #3B250E;--weui-ORANGE-80: #D39758;--weui-ORANGE-90: #CD8943;--weui-ORANGE-BG-100: #BB6000;--weui-ORANGE-BG-110: #A85600;--weui-ORANGE-BG-130: #824300;--weui-ORANGE-BG-90: #C1701A;--weui-ORANGERED-100: #FF6146;--weui-OVERLAY: rgba(0, 0, 0, .8);--weui-OVERLAY-WHITE: rgba(242, 242, 242, .8);--weui-PURPLE-100: #8183FF;--weui-PURPLE-120: #6768CC;--weui-PURPLE-170: #26274C;--weui-PURPLE-80: #9A9BFF;--weui-PURPLE-90: #8D8FFF;--weui-PURPLE-BG-100: #6768CC;--weui-PURPLE-BG-110: #5C5DB7;--weui-PURPLE-BG-130: #48498F;--weui-PURPLE-BG-90: #7677D1;--weui-RED-100: #FA5151;--weui-RED-120: #C84040;--weui-RED-170: #4B1818;--weui-RED-80: #FB7373;--weui-RED-90: #FA6262;--weui-RED-BG-100: #CF5148;--weui-RED-BG-110: #BA4940;--weui-RED-BG-130: #913832;--weui-RED-BG-90: #D3625A;--weui-SECONDARY-BG: rgba(255, 255, 255, .1);--weui-SEPARATOR-0: rgba(255, 255, 255, .05);--weui-SEPARATOR-1: rgba(255, 255, 255, .15);--weui-STATELAYER-HOVERED: rgba(0, 0, 0, .02);--weui-STATELAYER-PRESSED: rgba(255, 255, 255, .1);--weui-STATELAYER-PRESSEDSTRENGTHENED: rgba(255, 255, 255, .2);--weui-YELLOW-100: #CC9C00;--weui-YELLOW-120: #A37C00;--weui-YELLOW-170: #3D2F00;--weui-YELLOW-80: #D6AF33;--weui-YELLOW-90: #D1A519;--weui-YELLOW-BG-100: #BF9100;--weui-YELLOW-BG-110: #AB8200;--weui-YELLOW-BG-130: #866500;--weui-YELLOW-BG-90: #C59C1A;--weui-FG-HALF: rgba(255, 255, 255, .6);--weui-RED: #FA5151;--weui-ORANGERED: #FF6146;--weui-ORANGE: #C87D2F;--weui-YELLOW: #CC9C00;--weui-GREEN: #74A800;--weui-LIGHTGREEN: #3EB575;--weui-TEXTGREEN: #259C5C;--weui-BRAND: #07C160;--weui-BLUE: #10AEFF;--weui-INDIGO: #1196FF;--weui-PURPLE: #8183FF;--weui-LINK: #7D90A9;--weui-REDORANGE: #FF6146;--weui-TAG-TEXT-BLACK: rgba(255, 255, 255, .5);--weui-TAG-BACKGROUND-BLACK: rgba(255, 255, 255, .05);--weui-WHITE: rgba(255, 255, 255, .8);--weui-FG: #fff;--weui-BG: #000;--weui-FG-5: rgba(255, 255, 255, .1);--weui-TAG-BACKGROUND-ORANGE: rgba(250, 157, 59, .1);--weui-TAG-BACKGROUND-GREEN: rgba(6, 174, 86, .1);--weui-TAG-TEXT-RED: rgba(250, 81, 81, .6);--weui-TAG-BACKGROUND-RED: rgba(250, 81, 81, .1);--weui-TAG-BACKGROUND-BLUE: rgba(16, 174, 255, .1);--weui-TAG-TEXT-ORANGE: rgba(250, 157, 59, .6);--weui-TAG-TEXT-GREEN: rgba(6, 174, 86, .6);--weui-TAG-TEXT-BLUE: rgba(16, 174, 255, .6)}.rich_media_content{color:#000000e5;font-size:var(--articleFontsize);overflow:hidden;text-align:justify}.rich_media_content{color:var(--weui-FG-HALF)}.rich_media_content{position:relative;z-index:0}@media(prefers-color-scheme:dark){:root{--win-scrollbar-hover-bgcolor: rgba(255, 255, 255, .55);--win-scrollbar-bgcolor: rgba(255, 255, 255, .5)}}@media screen and (min-width:1024px){body:not(.pages_skin_pc) :root{--appmsgPageGap: 20px;--appmsgPageBottomGap: 40px}}:root{--articleFontsize: 17px}:root{--sab: env(safe-area-inset-bottom)}:root{--win-scrollbar-hover-bgcolor: rgba(0, 0, 0, .55);--win-scrollbar-bgcolor: rgba(0, 0, 0, .5)}:root{--wxBorderAvatarRatio: 3}:root{--discussPageGap: 20px}:root{--immersive-safe-bottom: env(safe-area-inset-bottom)}:root{--immersive-safe-bottom: env(safe-area-inset-bottom);--pc-drawer-comment-topbar-height: 40px}:root{--appmsgPageGap: 20px;--appmsgPageBottomGap: 40px}.rich_media_content p{clear:both;min-height:1em}td p{margin:0;padding:0}*{margin:0;padding:0}.rich_media_content *{max-width:100%!important;box-sizing:border-box!important;-webkit-box-sizing:border-box!important;word-wrap:break-word!important}a{color:#576b95;text-decoration:none;-webkit-tap-highlight-color:#0000;-webkit-user-drag:none}a{color:var(--weui-LINK)}a{cursor:default}body,.wx-root,page{--weui-BTN-HEIGHT: 48;--weui-BTN-HEIGHT-MEDIUM: 40;--weui-BTN-HEIGHT-SMALL: 32}@media(prefers-color-scheme:dark){body:not([data-weui-theme=light]){--weui-elpsColor: rgba(255, 255, 255, .8)}}@media(prefers-color-scheme:dark){body:not([data-weui-theme=light]){--weui-mask-elpsColor: rgba(255, 255, 255, .8);--weui-mask-gradient: linear-gradient(to right, rgba(25, 25, 25, 0), #191919 40%)}}@media screen and (min-width:1024px){body:not(.pages_skin_pc) body,body:not(.pages_skin_pc) .wx-root{--weui-BG-0: #EDEDED;--weui-BG-1: #F7F7F7;--weui-BG-2: #FFFFFF;--weui-BG-3: #F7F7F7;--weui-BG-4: #4C4C4C;--weui-BG-5: #FFFFFF;--weui-BLUE-100: #10AEFF;--weui-BLUE-120: #3FBEFF;--weui-BLUE-170: #B7E6FF;--weui-BLUE-80: #0C8BCC;--weui-BLUE-90: #0E9CE6;--weui-BLUE-BG-100: #48A6E2;--weui-BLUE-BG-110: #5AAFE4;--weui-BLUE-BG-130: #7FC0EA;--weui-BLUE-BG-90: #4095CB;--weui-BRAND-100: #07C160;--weui-BRAND-120: #38CD7F;--weui-BRAND-170: #B4ECCE;--weui-BRAND-80: #059A4C;--weui-BRAND-90: #06AE56;--weui-BRAND-BG-100: #2AAE67;--weui-BRAND-BG-110: #3EB575;--weui-BRAND-BG-130: #69C694;--weui-BRAND-BG-90: #259C5C;--weui-FG-0: rgba(0, 0, 0, .9);--weui-FG-0_5: rgba(0, 0, 0, .9);--weui-FG-1: rgba(0, 0, 0, .55);--weui-FG-2: rgba(0, 0, 0, .3);--weui-FG-3: rgba(0, 0, 0, .1);--weui-FG-4: rgba(0, 0, 0, .15);--weui-GLYPH-0: rgba(0, 0, 0, .9);--weui-GLYPH-1: rgba(0, 0, 0, .55);--weui-GLYPH-2: rgba(0, 0, 0, .3);--weui-GLYPH-WHITE-0: rgba(255, 255, 255, .8);--weui-GLYPH-WHITE-1: rgba(255, 255, 255, .5);--weui-GLYPH-WHITE-2: rgba(255, 255, 255, .3);--weui-GLYPH-WHITE-3: #FFFFFF;--weui-GREEN-100: #91D300;--weui-GREEN-120: #A7DB33;--weui-GREEN-170: #DEF1B3;--weui-GREEN-80: #74A800;--weui-GREEN-90: #82BD00;--weui-GREEN-BG-100: #96BE40;--weui-GREEN-BG-110: #A0C452;--weui-GREEN-BG-130: #B5D179;--weui-GREEN-BG-90: #86AA39;--weui-INDIGO-100: #1485EE;--weui-INDIGO-120: #439DF1;--weui-INDIGO-170: #B8DAF9;--weui-INDIGO-80: #106ABE;--weui-INDIGO-90: #1277D6;--weui-INDIGO-BG-100: #2B77BF;--weui-INDIGO-BG-110: #3F84C5;--weui-INDIGO-BG-130: #6BA0D2;--weui-INDIGO-BG-90: #266AAB;--weui-LIGHTGREEN-100: #95EC69;--weui-LIGHTGREEN-120: #AAEF87;--weui-LIGHTGREEN-170: #DEF9D1;--weui-LIGHTGREEN-80: #77BC54;--weui-LIGHTGREEN-90: #85D35E;--weui-LIGHTGREEN-BG-100: #72CF60;--weui-LIGHTGREEN-BG-110: #80D370;--weui-LIGHTGREEN-BG-130: #9CDD90;--weui-LIGHTGREEN-BG-90: #66B956;--weui-LINK-100: #576B95;--weui-LINK-120: #7888AA;--weui-LINK-170: #CCD2DE;--weui-LINK-80: #455577;--weui-LINK-90: #4E6085;--weui-LINKFINDER-100: #002666;--weui-MATERIAL-ATTACHMENTCOLUMN: rgba(245, 245, 245, .95);--weui-MATERIAL-NAVIGATIONBAR: rgba(237, 237, 237, .94);--weui-MATERIAL-REGULAR: rgba(247, 247, 247, .3);--weui-MATERIAL-THICK: rgba(247, 247, 247, .8);--weui-MATERIAL-THIN: rgba(255, 255, 255, .2);--weui-MATERIAL-TOOLBAR: rgba(246, 246, 246, .82);--weui-ORANGE-100: #FA9D3B;--weui-ORANGE-120: #FBB062;--weui-ORANGE-170: #FDE1C3;--weui-ORANGE-80: #C87D2F;--weui-ORANGE-90: #E08C34;--weui-ORANGE-BG-100: #EA7800;--weui-ORANGE-BG-110: #EC8519;--weui-ORANGE-BG-130: #F0A04D;--weui-ORANGE-BG-90: #D26B00;--weui-ORANGERED-100: #FF6146;--weui-OVERLAY: rgba(0, 0, 0, .5);--weui-OVERLAY-WHITE: rgba(242, 242, 242, .8);--weui-PURPLE-100: #6467F0;--weui-PURPLE-120: #8385F3;--weui-PURPLE-170: #D0D1FA;--weui-PURPLE-80: #5052C0;--weui-PURPLE-90: #595CD7;--weui-PURPLE-BG-100: #6769BA;--weui-PURPLE-BG-110: #7678C1;--weui-PURPLE-BG-130: #9496CE;--weui-PURPLE-BG-90: #5C5EA7;--weui-RED-100: #FA5151;--weui-RED-120: #FB7373;--weui-RED-170: #FDCACA;--weui-RED-80: #C84040;--weui-RED-90: #E14949;--weui-RED-BG-100: #CF5148;--weui-RED-BG-110: #D3625A;--weui-RED-BG-130: #DD847E;--weui-RED-BG-90: #B94840;--weui-SECONDARY-BG: rgba(0, 0, 0, .05);--weui-SEPARATOR-0: rgba(0, 0, 0, .1);--weui-SEPARATOR-1: rgba(0, 0, 0, .15);--weui-STATELAYER-HOVERED: rgba(0, 0, 0, .02);--weui-STATELAYER-PRESSED: rgba(0, 0, 0, .1);--weui-STATELAYER-PRESSEDSTRENGTHENED: rgba(0, 0, 0, .2);--weui-YELLOW-100: #FFC300;--weui-YELLOW-120: #FFCF33;--weui-YELLOW-170: #FFECB2;--weui-YELLOW-80: #CC9C00;--weui-YELLOW-90: #E6AF00;--weui-YELLOW-BG-100: #EFB600;--weui-YELLOW-BG-110: #F0BD19;--weui-YELLOW-BG-130: #F3CC4D;--weui-YELLOW-BG-90: #D7A400;--weui-FG-HALF: rgba(0, 0, 0, .9);--weui-RED: #FA5151;--weui-ORANGERED: #FF6146;--weui-ORANGE: #FA9D3B;--weui-YELLOW: #FFC300;--weui-GREEN: #91D300;--weui-LIGHTGREEN: #95EC69;--weui-TEXTGREEN: #06AE56;--weui-BRAND: #07C160;--weui-BLUE: #10AEFF;--weui-INDIGO: #1485EE;--weui-PURPLE: #6467F0;--weui-LINK: #576B95;--weui-TAG-TEXT-ORANGE: #FA9D3B;--weui-TAG-TEXT-GREEN: #06AE56;--weui-TAG-TEXT-BLUE: #10AEFF;--weui-REDORANGE: #FF6146;--weui-TAG-TEXT-BLACK: rgba(0, 0, 0, .5);--weui-TAG-BACKGROUND-BLACK: rgba(0, 0, 0, .05);--weui-WHITE: #FFFFFF;--weui-BG: #FFFFFF;--weui-FG: #000;--weui-FG-5: rgba(0, 0, 0, .05);--weui-TAG-BACKGROUND-ORANGE: rgba(250, 157, 59, .1);--weui-TAG-BACKGROUND-GREEN: rgba(6, 174, 86, .1);--weui-TAG-TEXT-RED: rgba(250, 81, 81, .6);--weui-TAG-BACKGROUND-RED: rgba(250, 81, 81, .1);--weui-TAG-BACKGROUND-BLUE: rgba(16, 174, 255, .1)}}@media screen and (min-width:1024px){body:not(.pages_skin_pc) .wx-root[data-weui-mode=care],body:not(.pages_skin_pc) body[data-weui-mode=care]{--weui-BG-0: #EDEDED;--weui-BG-1: #F7F7F7;--weui-BG-2: #FFFFFF;--weui-BG-3: #F7F7F7;--weui-BG-4: #4C4C4C;--weui-BG-5: #FFFFFF;--weui-BLUE-100: #007DBB;--weui-BLUE-120: #3FBEFF;--weui-BLUE-170: #B7E6FF;--weui-BLUE-80: #0C8BCC;--weui-BLUE-90: #0E9CE6;--weui-BLUE-BG-100: #48A6E2;--weui-BLUE-BG-110: #5AAFE4;--weui-BLUE-BG-130: #7FC0EA;--weui-BLUE-BG-90: #4095CB;--weui-BRAND-100: #018942;--weui-BRAND-120: #38CD7F;--weui-BRAND-170: #B4ECCE;--weui-BRAND-80: #059A4C;--weui-BRAND-90: #06AE56;--weui-BRAND-BG-100: #2AAE67;--weui-BRAND-BG-110: #3EB575;--weui-BRAND-BG-130: #69C694;--weui-BRAND-BG-90: #259C5C;--weui-FG-0: #000000;--weui-FG-0_5: #000000;--weui-FG-1: rgba(0, 0, 0, .6);--weui-FG-2: rgba(0, 0, 0, .42);--weui-FG-3: rgba(0, 0, 0, .1);--weui-FG-4: rgba(0, 0, 0, .15);--weui-GLYPH-0: #000000;--weui-GLYPH-1: rgba(0, 0, 0, .6);--weui-GLYPH-2: rgba(0, 0, 0, .42);--weui-GLYPH-WHITE-0: rgba(255, 255, 255, .85);--weui-GLYPH-WHITE-1: rgba(255, 255, 255, .55);--weui-GLYPH-WHITE-2: rgba(255, 255, 255, .35);--weui-GLYPH-WHITE-3: #FFFFFF;--weui-GREEN-100: #4F8400;--weui-GREEN-120: #A7DB33;--weui-GREEN-170: #DEF1B3;--weui-GREEN-80: #74A800;--weui-GREEN-90: #82BD00;--weui-GREEN-BG-100: #96BE40;--weui-GREEN-BG-110: #A0C452;--weui-GREEN-BG-130: #B5D179;--weui-GREEN-BG-90: #86AA39;--weui-INDIGO-100: #0075E2;--weui-INDIGO-120: #439DF1;--weui-INDIGO-170: #B8DAF9;--weui-INDIGO-80: #106ABE;--weui-INDIGO-90: #1277D6;--weui-INDIGO-BG-100: #2B77BF;--weui-INDIGO-BG-110: #3F84C5;--weui-INDIGO-BG-130: #6BA0D2;--weui-INDIGO-BG-90: #266AAB;--weui-LIGHTGREEN-100: #2E8800;--weui-LIGHTGREEN-120: #AAEF87;--weui-LIGHTGREEN-170: #DEF9D1;--weui-LIGHTGREEN-80: #77BC54;--weui-LIGHTGREEN-90: #85D35E;--weui-LIGHTGREEN-BG-100: #72CF60;--weui-LIGHTGREEN-BG-110: #80D370;--weui-LIGHTGREEN-BG-130: #9CDD90;--weui-LIGHTGREEN-BG-90: #66B956;--weui-LINK-100: #576B95;--weui-LINK-120: #7888AA;--weui-LINK-170: #CCD2DE;--weui-LINK-80: #455577;--weui-LINK-90: #4E6085;--weui-LINKFINDER-100: #002666;--weui-MATERIAL-ATTACHMENTCOLUMN: rgba(245, 245, 245, .95);--weui-MATERIAL-NAVIGATIONBAR: rgba(237, 237, 237, .94);--weui-MATERIAL-REGULAR: rgba(247, 247, 247, .3);--weui-MATERIAL-THICK: rgba(247, 247, 247, .8);--weui-MATERIAL-THIN: rgba(255, 255, 255, .2);--weui-MATERIAL-TOOLBAR: rgba(246, 246, 246, .82);--weui-ORANGE-100: #E17719;--weui-ORANGE-120: #FBB062;--weui-ORANGE-170: #FDE1C3;--weui-ORANGE-80: #C87D2F;--weui-ORANGE-90: #E08C34;--weui-ORANGE-BG-100: #EA7800;--weui-ORANGE-BG-110: #EC8519;--weui-ORANGE-BG-130: #F0A04D;--weui-ORANGE-BG-90: #D26B00;--weui-ORANGERED-100: #D14730;--weui-OVERLAY: rgba(0, 0, 0, .5);--weui-OVERLAY-WHITE: rgba(242, 242, 242, .8);--weui-PURPLE-100: #6265F1;--weui-PURPLE-120: #8385F3;--weui-PURPLE-170: #D0D1FA;--weui-PURPLE-80: #5052C0;--weui-PURPLE-90: #595CD7;--weui-PURPLE-BG-100: #6769BA;--weui-PURPLE-BG-110: #7678C1;--weui-PURPLE-BG-130: #9496CE;--weui-PURPLE-BG-90: #5C5EA7;--weui-RED-100: #DC3636;--weui-RED-120: #FB7373;--weui-RED-170: #FDCACA;--weui-RED-80: #C84040;--weui-RED-90: #E14949;--weui-RED-BG-100: #CF5148;--weui-RED-BG-110: #D3625A;--weui-RED-BG-130: #DD847E;--weui-RED-BG-90: #B94840;--weui-SECONDARY-BG: rgba(0, 0, 0, .1);--weui-SEPARATOR-0: rgba(0, 0, 0, .1);--weui-SEPARATOR-1: rgba(0, 0, 0, .15);--weui-STATELAYER-HOVERED: rgba(0, 0, 0, .02);--weui-STATELAYER-PRESSED: rgba(0, 0, 0, .1);--weui-STATELAYER-PRESSEDSTRENGTHENED: rgba(0, 0, 0, .2);--weui-YELLOW-100: #BB8E00;--weui-YELLOW-120: #FFCF33;--weui-YELLOW-170: #FFECB2;--weui-YELLOW-80: #CC9C00;--weui-YELLOW-90: #E6AF00;--weui-YELLOW-BG-100: #EFB600;--weui-YELLOW-BG-110: #F0BD19;--weui-YELLOW-BG-130: #F3CC4D;--weui-YELLOW-BG-90: #D7A400;--weui-FG-HALF: #000000;--weui-RED: #DC3636;--weui-ORANGERED: #D14730;--weui-ORANGE: #E17719;--weui-YELLOW: #BB8E00;--weui-GREEN: #4F8400;--weui-LIGHTGREEN: #2E8800;--weui-TEXTGREEN: #06AE56;--weui-BRAND: #018942;--weui-BLUE: #007DBB;--weui-INDIGO: #0075E2;--weui-PURPLE: #6265F1;--weui-LINK: #576B95;--weui-TAG-TEXT-ORANGE: #E17719;--weui-TAG-TEXT-GREEN: #06AE56;--weui-TAG-TEXT-BLUE: #007DBB;--weui-REDORANGE: #D14730;--weui-TAG-TEXT-BLACK: rgba(0, 0, 0, .5);--weui-WHITE: #FFFFFF;--weui-BG: #FFFFFF;--weui-FG: #000;--weui-FG-5: rgba(0, 0, 0, .05);--weui-TAG-BACKGROUND-ORANGE: rgba(225, 119, 25, .1);--weui-TAG-BACKGROUND-GREEN: rgba(6, 174, 86, .1);--weui-TAG-TEXT-RED: rgba(250, 81, 81, .6);--weui-TAG-BACKGROUND-RED: rgba(250, 81, 81, .1);--weui-TAG-BACKGROUND-BLUE: rgba(0, 125, 187, .1);--weui-TAG-BACKGROUND-BLACK: rgba(0, 0, 0, .05)}}@media screen and (min-width:1024px){body:not(.pages_skin_pc) .wx-root[data-weui-mode=care][data-weui-theme=dark],body:not(.pages_skin_pc) body[data-weui-mode=care][data-weui-theme=dark]{--weui-BG-0: #111111;--weui-BG-1: #1E1E1E;--weui-BG-2: #191919;--weui-BG-3: #202020;--weui-BG-4: #404040;--weui-BG-5: #2C2C2C;--weui-BLUE-100: #10AEFF;--weui-BLUE-120: #0C8BCC;--weui-BLUE-170: #04344D;--weui-BLUE-80: #3FBEFF;--weui-BLUE-90: #28B6FF;--weui-BLUE-BG-100: #48A6E2;--weui-BLUE-BG-110: #4095CB;--weui-BLUE-BG-130: #32749E;--weui-BLUE-BG-90: #5AAFE4;--weui-BRAND-100: #07C160;--weui-BRAND-120: #059A4C;--weui-BRAND-170: #023A1C;--weui-BRAND-80: #38CD7F;--weui-BRAND-90: #20C770;--weui-BRAND-BG-100: #2AAE67;--weui-BRAND-BG-110: #259C5C;--weui-BRAND-BG-130: #1D7A48;--weui-BRAND-BG-90: #3EB575;--weui-FG-0: rgba(255, 255, 255, .85);--weui-FG-0_5: rgba(255, 255, 255, .65);--weui-FG-1: rgba(255, 255, 255, .55);--weui-FG-2: rgba(255, 255, 255, .35);--weui-FG-3: rgba(255, 255, 255, .1);--weui-FG-4: rgba(255, 255, 255, .15);--weui-GLYPH-0: rgba(255, 255, 255, .85);--weui-GLYPH-1: rgba(255, 255, 255, .55);--weui-GLYPH-2: rgba(255, 255, 255, .35);--weui-GLYPH-WHITE-0: rgba(255, 255, 255, .85);--weui-GLYPH-WHITE-1: rgba(255, 255, 255, .55);--weui-GLYPH-WHITE-2: rgba(255, 255, 255, .35);--weui-GLYPH-WHITE-3: #FFFFFF;--weui-GREEN-100: #74A800;--weui-GREEN-120: #5C8600;--weui-GREEN-170: #233200;--weui-GREEN-80: #8FB933;--weui-GREEN-90: #82B01A;--weui-GREEN-BG-100: #789833;--weui-GREEN-BG-110: #6B882D;--weui-GREEN-BG-130: #65802B;--weui-GREEN-BG-90: #85A247;--weui-INDIGO-100: #1196FF;--weui-INDIGO-120: #0D78CC;--weui-INDIGO-170: #052D4D;--weui-INDIGO-80: #40ABFF;--weui-INDIGO-90: #28A0FF;--weui-INDIGO-BG-100: #0D78CC;--weui-INDIGO-BG-110: #0B6BB7;--weui-INDIGO-BG-130: #09548F;--weui-INDIGO-BG-90: #2585D1;--weui-LIGHTGREEN-100: #3EB575;--weui-LIGHTGREEN-120: #31905D;--weui-LIGHTGREEN-170: #123522;--weui-LIGHTGREEN-80: #64C390;--weui-LIGHTGREEN-90: #51BC83;--weui-LIGHTGREEN-BG-100: #31905D;--weui-LIGHTGREEN-BG-110: #2C8153;--weui-LIGHTGREEN-BG-130: #226541;--weui-LIGHTGREEN-BG-90: #31905D;--weui-LINK-100: #7D90A9;--weui-LINK-120: #647387;--weui-LINK-170: #252A32;--weui-LINK-80: #97A6BA;--weui-LINK-90: #899AB1;--weui-LINKFINDER-100: #DEE9FF;--weui-MATERIAL-ATTACHMENTCOLUMN: rgba(32, 32, 32, .93);--weui-MATERIAL-NAVIGATIONBAR: rgba(18, 18, 18, .9);--weui-MATERIAL-REGULAR: rgba(37, 37, 37, .6);--weui-MATERIAL-THICK: rgba(34, 34, 34, .9);--weui-MATERIAL-THIN: rgba(245, 245, 245, .4);--weui-MATERIAL-TOOLBAR: rgba(35, 35, 35, .93);--weui-ORANGE-100: #C87D2F;--weui-ORANGE-120: #A06425;--weui-ORANGE-170: #3B250E;--weui-ORANGE-80: #D39758;--weui-ORANGE-90: #CD8943;--weui-ORANGE-BG-100: #BB6000;--weui-ORANGE-BG-110: #A85600;--weui-ORANGE-BG-130: #824300;--weui-ORANGE-BG-90: #C1701A;--weui-ORANGERED-100: #FF6146;--weui-OVERLAY: rgba(0, 0, 0, .8);--weui-OVERLAY-WHITE: rgba(242, 242, 242, .8);--weui-PURPLE-100: #8183FF;--weui-PURPLE-120: #6768CC;--weui-PURPLE-170: #26274C;--weui-PURPLE-80: #9A9BFF;--weui-PURPLE-90: #8D8FFF;--weui-PURPLE-BG-100: #6768CC;--weui-PURPLE-BG-110: #5C5DB7;--weui-PURPLE-BG-130: #48498F;--weui-PURPLE-BG-90: #7677D1;--weui-RED-100: #FA5151;--weui-RED-120: #C84040;--weui-RED-170: #4B1818;--weui-RED-80: #FB7373;--weui-RED-90: #FA6262;--weui-RED-BG-100: #CF5148;--weui-RED-BG-110: #BA4940;--weui-RED-BG-130: #913832;--weui-RED-BG-90: #D3625A;--weui-SECONDARY-BG: rgba(255, 255, 255, .15);--weui-SEPARATOR-0: rgba(255, 255, 255, .05);--weui-SEPARATOR-1: rgba(255, 255, 255, .15);--weui-STATELAYER-HOVERED: rgba(0, 0, 0, .02);--weui-STATELAYER-PRESSED: rgba(255, 255, 255, .1);--weui-STATELAYER-PRESSEDSTRENGTHENED: rgba(255, 255, 255, .2);--weui-YELLOW-100: #CC9C00;--weui-YELLOW-120: #A37C00;--weui-YELLOW-170: #3D2F00;--weui-YELLOW-80: #D6AF33;--weui-YELLOW-90: #D1A519;--weui-YELLOW-BG-100: #BF9100;--weui-YELLOW-BG-110: #AB8200;--weui-YELLOW-BG-130: #866500;--weui-YELLOW-BG-90: #C59C1A;--weui-FG-HALF: rgba(255, 255, 255, .65);--weui-RED: #FA5151;--weui-ORANGERED: #FF6146;--weui-ORANGE: #C87D2F;--weui-YELLOW: #CC9C00;--weui-GREEN: #74A800;--weui-LIGHTGREEN: #3EB575;--weui-TEXTGREEN: #259C5C;--weui-BRAND: #07C160;--weui-BLUE: #10AEFF;--weui-INDIGO: #1196FF;--weui-PURPLE: #8183FF;--weui-LINK: #7D90A9;--weui-REDORANGE: #FF6146;--weui-TAG-BACKGROUND-BLACK: rgba(255, 255, 255, .05);--weui-FG: #fff;--weui-WHITE: rgba(255, 255, 255, .8);--weui-FG-5: rgba(255, 255, 255, .1);--weui-TAG-BACKGROUND-ORANGE: rgba(250, 157, 59, .1);--weui-TAG-BACKGROUND-GREEN: rgba(6, 174, 86, .1);--weui-TAG-TEXT-RED: rgba(250, 81, 81, .6);--weui-TAG-BACKGROUND-RED: rgba(250, 81, 81, .1);--weui-TAG-BACKGROUND-BLUE: rgba(16, 174, 255, .1);--weui-TAG-TEXT-ORANGE: rgba(250, 157, 59, .6);--weui-BG: #000;--weui-TAG-TEXT-GREEN: rgba(6, 174, 86, .6);--weui-TAG-TEXT-BLUE: rgba(16, 174, 255, .6);--weui-TAG-TEXT-BLACK: rgba(255, 255, 255, .5)}}@media screen and (min-width:1024px){body:not(.pages_skin_pc) .wx-root,body:not(.pages_skin_pc) body{--appmsgExtra-BG: #F7F7F7}}@media screen and (min-width:1024px){body:not(.pages_skin_pc) .wx-root[data-weui-theme=dark],body:not(.pages_skin_pc) body[data-weui-theme=dark]{--weui-BG-0: #111111;--weui-BG-1: #1E1E1E;--weui-BG-2: #191919;--weui-BG-3: #202020;--weui-BG-4: #404040;--weui-BG-5: #2C2C2C;--weui-BLUE-100: #10AEFF;--weui-BLUE-120: #0C8BCC;--weui-BLUE-170: #04344D;--weui-BLUE-80: #3FBEFF;--weui-BLUE-90: #28B6FF;--weui-BLUE-BG-100: #48A6E2;--weui-BLUE-BG-110: #4095CB;--weui-BLUE-BG-130: #32749E;--weui-BLUE-BG-90: #5AAFE4;--weui-BRAND-100: #07C160;--weui-BRAND-120: #059A4C;--weui-BRAND-170: #023A1C;--weui-BRAND-80: #38CD7F;--weui-BRAND-90: #20C770;--weui-BRAND-BG-100: #2AAE67;--weui-BRAND-BG-110: #259C5C;--weui-BRAND-BG-130: #1D7A48;--weui-BRAND-BG-90: #3EB575;--weui-FG-0: rgba(255, 255, 255, .8);--weui-FG-0_5: rgba(255, 255, 255, .6);--weui-FG-1: rgba(255, 255, 255, .5);--weui-FG-2: rgba(255, 255, 255, .3);--weui-FG-3: rgba(255, 255, 255, .1);--weui-FG-4: rgba(255, 255, 255, .15);--weui-GLYPH-0: rgba(255, 255, 255, .8);--weui-GLYPH-1: rgba(255, 255, 255, .5);--weui-GLYPH-2: rgba(255, 255, 255, .3);--weui-GLYPH-WHITE-0: rgba(255, 255, 255, .8);--weui-GLYPH-WHITE-1: rgba(255, 255, 255, .5);--weui-GLYPH-WHITE-2: rgba(255, 255, 255, .3);--weui-GLYPH-WHITE-3: #FFFFFF;--weui-GREEN-100: #74A800;--weui-GREEN-120: #5C8600;--weui-GREEN-170: #233200;--weui-GREEN-80: #8FB933;--weui-GREEN-90: #82B01A;--weui-GREEN-BG-100: #789833;--weui-GREEN-BG-110: #6B882D;--weui-GREEN-BG-130: #65802B;--weui-GREEN-BG-90: #85A247;--weui-INDIGO-100: #1196FF;--weui-INDIGO-120: #0D78CC;--weui-INDIGO-170: #052D4D;--weui-INDIGO-80: #40ABFF;--weui-INDIGO-90: #28A0FF;--weui-INDIGO-BG-100: #0D78CC;--weui-INDIGO-BG-110: #0B6BB7;--weui-INDIGO-BG-130: #09548F;--weui-INDIGO-BG-90: #2585D1;--weui-LIGHTGREEN-100: #3EB575;--weui-LIGHTGREEN-120: #31905D;--weui-LIGHTGREEN-170: #123522;--weui-LIGHTGREEN-80: #64C390;--weui-LIGHTGREEN-90: #51BC83;--weui-LIGHTGREEN-BG-100: #31905D;--weui-LIGHTGREEN-BG-110: #2C8153;--weui-LIGHTGREEN-BG-130: #226541;--weui-LIGHTGREEN-BG-90: #31905D;--weui-LINK-100: #7D90A9;--weui-LINK-120: #647387;--weui-LINK-170: #252A32;--weui-LINK-80: #97A6BA;--weui-LINK-90: #899AB1;--weui-LINKFINDER-100: #DEE9FF;--weui-MATERIAL-ATTACHMENTCOLUMN: rgba(32, 32, 32, .93);--weui-MATERIAL-NAVIGATIONBAR: rgba(18, 18, 18, .9);--weui-MATERIAL-REGULAR: rgba(37, 37, 37, .6);--weui-MATERIAL-THICK: rgba(34, 34, 34, .9);--weui-MATERIAL-THIN: rgba(95, 95, 95, .4);--weui-MATERIAL-TOOLBAR: rgba(35, 35, 35, .93);--weui-ORANGE-100: #C87D2F;--weui-ORANGE-120: #A06425;--weui-ORANGE-170: #3B250E;--weui-ORANGE-80: #D39758;--weui-ORANGE-90: #CD8943;--weui-ORANGE-BG-100: #BB6000;--weui-ORANGE-BG-110: #A85600;--weui-ORANGE-BG-130: #824300;--weui-ORANGE-BG-90: #C1701A;--weui-ORANGERED-100: #FF6146;--weui-OVERLAY: rgba(0, 0, 0, .8);--weui-OVERLAY-WHITE: rgba(242, 242, 242, .8);--weui-PURPLE-100: #8183FF;--weui-PURPLE-120: #6768CC;--weui-PURPLE-170: #26274C;--weui-PURPLE-80: #9A9BFF;--weui-PURPLE-90: #8D8FFF;--weui-PURPLE-BG-100: #6768CC;--weui-PURPLE-BG-110: #5C5DB7;--weui-PURPLE-BG-130: #48498F;--weui-PURPLE-BG-90: #7677D1;--weui-RED-100: #FA5151;--weui-RED-120: #C84040;--weui-RED-170: #4B1818;--weui-RED-80: #FB7373;--weui-RED-90: #FA6262;--weui-RED-BG-100: #CF5148;--weui-RED-BG-110: #BA4940;--weui-RED-BG-130: #913832;--weui-RED-BG-90: #D3625A;--weui-SECONDARY-BG: rgba(255, 255, 255, .1);--weui-SEPARATOR-0: rgba(255, 255, 255, .05);--weui-SEPARATOR-1: rgba(255, 255, 255, .15);--weui-STATELAYER-HOVERED: rgba(0, 0, 0, .02);--weui-STATELAYER-PRESSED: rgba(255, 255, 255, .1);--weui-STATELAYER-PRESSEDSTRENGTHENED: rgba(255, 255, 255, .2);--weui-YELLOW-100: #CC9C00;--weui-YELLOW-120: #A37C00;--weui-YELLOW-170: #3D2F00;--weui-YELLOW-80: #D6AF33;--weui-YELLOW-90: #D1A519;--weui-YELLOW-BG-100: #BF9100;--weui-YELLOW-BG-110: #AB8200;--weui-YELLOW-BG-130: #866500;--weui-YELLOW-BG-90: #C59C1A;--weui-FG-HALF: rgba(255, 255, 255, .6);--weui-RED: #FA5151;--weui-ORANGERED: #FF6146;--weui-ORANGE: #C87D2F;--weui-YELLOW: #CC9C00;--weui-GREEN: #74A800;--weui-LIGHTGREEN: #3EB575;--weui-TEXTGREEN: #259C5C;--weui-BRAND: #07C160;--weui-BLUE: #10AEFF;--weui-INDIGO: #1196FF;--weui-PURPLE: #8183FF;--weui-LINK: #7D90A9;--weui-REDORANGE: #FF6146;--weui-TAG-TEXT-BLACK: rgba(255, 255, 255, .5);--weui-TAG-BACKGROUND-BLACK: rgba(255, 255, 255, .05);--weui-WHITE: rgba(255, 255, 255, .8);--weui-FG: #fff;--weui-BG: #000;--weui-FG-5: rgba(255, 255, 255, .1);--weui-TAG-BACKGROUND-ORANGE: rgba(250, 157, 59, .1);--weui-TAG-BACKGROUND-GREEN: rgba(6, 174, 86, .1);--weui-TAG-TEXT-RED: rgba(250, 81, 81, .6);--weui-TAG-BACKGROUND-RED: rgba(250, 81, 81, .1);--weui-TAG-BACKGROUND-BLUE: rgba(16, 174, 255, .1);--weui-TAG-TEXT-ORANGE: rgba(250, 157, 59, .6);--weui-TAG-TEXT-GREEN: rgba(6, 174, 86, .6);--weui-TAG-TEXT-BLUE: rgba(16, 174, 255, .6)}}@media screen and (min-width:1024px){body:not(.pages_skin_pc) .wx-root[data-weui-theme=dark],body:not(.pages_skin_pc) body[data-weui-theme=dark]{--appmsgExtra-BG: #121212}}@media screen and (min-width:1024px){body:not(.pages_skin_pc)body:not(.pages_skin_pc){background:var(--weui-BG-2)}}@media screen and (min-width:1024px) and (prefers-color-scheme:dark){body:not(.pages_skin_pc) .wx-root[data-weui-mode=care]:not([data-weui-theme=light]),body:not(.pages_skin_pc) body[data-weui-mode=care]:not([data-weui-theme=light]){--weui-BG-0: #111111;--weui-BG-1: #1E1E1E;--weui-BG-2: #191919;--weui-BG-3: #202020;--weui-BG-4: #404040;--weui-BG-5: #2C2C2C;--weui-BLUE-100: #10AEFF;--weui-BLUE-120: #0C8BCC;--weui-BLUE-170: #04344D;--weui-BLUE-80: #3FBEFF;--weui-BLUE-90: #28B6FF;--weui-BLUE-BG-100: #48A6E2;--weui-BLUE-BG-110: #4095CB;--weui-BLUE-BG-130: #32749E;--weui-BLUE-BG-90: #5AAFE4;--weui-BRAND-100: #07C160;--weui-BRAND-120: #059A4C;--weui-BRAND-170: #023A1C;--weui-BRAND-80: #38CD7F;--weui-BRAND-90: #20C770;--weui-BRAND-BG-100: #2AAE67;--weui-BRAND-BG-110: #259C5C;--weui-BRAND-BG-130: #1D7A48;--weui-BRAND-BG-90: #3EB575;--weui-FG-0: rgba(255, 255, 255, .85);--weui-FG-0_5: rgba(255, 255, 255, .65);--weui-FG-1: rgba(255, 255, 255, .55);--weui-FG-2: rgba(255, 255, 255, .35);--weui-FG-3: rgba(255, 255, 255, .1);--weui-FG-4: rgba(255, 255, 255, .15);--weui-GLYPH-0: rgba(255, 255, 255, .85);--weui-GLYPH-1: rgba(255, 255, 255, .55);--weui-GLYPH-2: rgba(255, 255, 255, .35);--weui-GLYPH-WHITE-0: rgba(255, 255, 255, .85);--weui-GLYPH-WHITE-1: rgba(255, 255, 255, .55);--weui-GLYPH-WHITE-2: rgba(255, 255, 255, .35);--weui-GLYPH-WHITE-3: #FFFFFF;--weui-GREEN-100: #74A800;--weui-GREEN-120: #5C8600;--weui-GREEN-170: #233200;--weui-GREEN-80: #8FB933;--weui-GREEN-90: #82B01A;--weui-GREEN-BG-100: #789833;--weui-GREEN-BG-110: #6B882D;--weui-GREEN-BG-130: #65802B;--weui-GREEN-BG-90: #85A247;--weui-INDIGO-100: #1196FF;--weui-INDIGO-120: #0D78CC;--weui-INDIGO-170: #052D4D;--weui-INDIGO-80: #40ABFF;--weui-INDIGO-90: #28A0FF;--weui-INDIGO-BG-100: #0D78CC;--weui-INDIGO-BG-110: #0B6BB7;--weui-INDIGO-BG-130: #09548F;--weui-INDIGO-BG-90: #2585D1;--weui-LIGHTGREEN-100: #3EB575;--weui-LIGHTGREEN-120: #31905D;--weui-LIGHTGREEN-170: #123522;--weui-LIGHTGREEN-80: #64C390;--weui-LIGHTGREEN-90: #51BC83;--weui-LIGHTGREEN-BG-100: #31905D;--weui-LIGHTGREEN-BG-110: #2C8153;--weui-LIGHTGREEN-BG-130: #226541;--weui-LIGHTGREEN-BG-90: #31905D;--weui-LINK-100: #7D90A9;--weui-LINK-120: #647387;--weui-LINK-170: #252A32;--weui-LINK-80: #97A6BA;--weui-LINK-90: #899AB1;--weui-LINKFINDER-100: #DEE9FF;--weui-MATERIAL-ATTACHMENTCOLUMN: rgba(32, 32, 32, .93);--weui-MATERIAL-NAVIGATIONBAR: rgba(18, 18, 18, .9);--weui-MATERIAL-REGULAR: rgba(37, 37, 37, .6);--weui-MATERIAL-THICK: rgba(34, 34, 34, .9);--weui-MATERIAL-THIN: rgba(245, 245, 245, .4);--weui-MATERIAL-TOOLBAR: rgba(35, 35, 35, .93);--weui-ORANGE-100: #C87D2F;--weui-ORANGE-120: #A06425;--weui-ORANGE-170: #3B250E;--weui-ORANGE-80: #D39758;--weui-ORANGE-90: #CD8943;--weui-ORANGE-BG-100: #BB6000;--weui-ORANGE-BG-110: #A85600;--weui-ORANGE-BG-130: #824300;--weui-ORANGE-BG-90: #C1701A;--weui-ORANGERED-100: #FF6146;--weui-OVERLAY: rgba(0, 0, 0, .8);--weui-OVERLAY-WHITE: rgba(242, 242, 242, .8);--weui-PURPLE-100: #8183FF;--weui-PURPLE-120: #6768CC;--weui-PURPLE-170: #26274C;--weui-PURPLE-80: #9A9BFF;--weui-PURPLE-90: #8D8FFF;--weui-PURPLE-BG-100: #6768CC;--weui-PURPLE-BG-110: #5C5DB7;--weui-PURPLE-BG-130: #48498F;--weui-PURPLE-BG-90: #7677D1;--weui-RED-100: #FA5151;--weui-RED-120: #C84040;--weui-RED-170: #4B1818;--weui-RED-80: #FB7373;--weui-RED-90: #FA6262;--weui-RED-BG-100: #CF5148;--weui-RED-BG-110: #BA4940;--weui-RED-BG-130: #913832;--weui-RED-BG-90: #D3625A;--weui-SECONDARY-BG: rgba(255, 255, 255, .15);--weui-SEPARATOR-0: rgba(255, 255, 255, .05);--weui-SEPARATOR-1: rgba(255, 255, 255, .15);--weui-STATELAYER-HOVERED: rgba(0, 0, 0, .02);--weui-STATELAYER-PRESSED: rgba(255, 255, 255, .1);--weui-STATELAYER-PRESSEDSTRENGTHENED: rgba(255, 255, 255, .2);--weui-YELLOW-100: #CC9C00;--weui-YELLOW-120: #A37C00;--weui-YELLOW-170: #3D2F00;--weui-YELLOW-80: #D6AF33;--weui-YELLOW-90: #D1A519;--weui-YELLOW-BG-100: #BF9100;--weui-YELLOW-BG-110: #AB8200;--weui-YELLOW-BG-130: #866500;--weui-YELLOW-BG-90: #C59C1A;--weui-FG-HALF: rgba(255, 255, 255, .65);--weui-RED: #FA5151;--weui-ORANGERED: #FF6146;--weui-ORANGE: #C87D2F;--weui-YELLOW: #CC9C00;--weui-GREEN: #74A800;--weui-LIGHTGREEN: #3EB575;--weui-TEXTGREEN: #259C5C;--weui-BRAND: #07C160;--weui-BLUE: #10AEFF;--weui-INDIGO: #1196FF;--weui-PURPLE: #8183FF;--weui-LINK: #7D90A9;--weui-REDORANGE: #FF6146;--weui-TAG-BACKGROUND-BLACK: rgba(255, 255, 255, .05);--weui-FG: #fff;--weui-WHITE: rgba(255, 255, 255, .8);--weui-FG-5: rgba(255, 255, 255, .1);--weui-TAG-BACKGROUND-ORANGE: rgba(250, 157, 59, .1);--weui-TAG-BACKGROUND-GREEN: rgba(6, 174, 86, .1);--weui-TAG-TEXT-RED: rgba(250, 81, 81, .6);--weui-TAG-BACKGROUND-RED: rgba(250, 81, 81, .1);--weui-TAG-BACKGROUND-BLUE: rgba(16, 174, 255, .1);--weui-TAG-TEXT-ORANGE: rgba(250, 157, 59, .6);--weui-BG: #000;--weui-TAG-TEXT-GREEN: rgba(6, 174, 86, .6);--weui-TAG-TEXT-BLUE: rgba(16, 174, 255, .6);--weui-TAG-TEXT-BLACK: rgba(255, 255, 255, .5)}}@media screen and (min-width:1024px) and (prefers-color-scheme:dark){body:not(.pages_skin_pc) .wx-root:not([data-weui-theme=light]),body:not(.pages_skin_pc) body:not([data-weui-theme=light]){--weui-BG-0: #111111;--weui-BG-1: #1E1E1E;--weui-BG-2: #191919;--weui-BG-3: #202020;--weui-BG-4: #404040;--weui-BG-5: #2C2C2C;--weui-BLUE-100: #10AEFF;--weui-BLUE-120: #0C8BCC;--weui-BLUE-170: #04344D;--weui-BLUE-80: #3FBEFF;--weui-BLUE-90: #28B6FF;--weui-BLUE-BG-100: #48A6E2;--weui-BLUE-BG-110: #4095CB;--weui-BLUE-BG-130: #32749E;--weui-BLUE-BG-90: #5AAFE4;--weui-BRAND-100: #07C160;--weui-BRAND-120: #059A4C;--weui-BRAND-170: #023A1C;--weui-BRAND-80: #38CD7F;--weui-BRAND-90: #20C770;--weui-BRAND-BG-100: #2AAE67;--weui-BRAND-BG-110: #259C5C;--weui-BRAND-BG-130: #1D7A48;--weui-BRAND-BG-90: #3EB575;--weui-FG-0: rgba(255, 255, 255, .8);--weui-FG-0_5: rgba(255, 255, 255, .6);--weui-FG-1: rgba(255, 255, 255, .5);--weui-FG-2: rgba(255, 255, 255, .3);--weui-FG-3: rgba(255, 255, 255, .1);--weui-FG-4: rgba(255, 255, 255, .15);--weui-GLYPH-0: rgba(255, 255, 255, .8);--weui-GLYPH-1: rgba(255, 255, 255, .5);--weui-GLYPH-2: rgba(255, 255, 255, .3);--weui-GLYPH-WHITE-0: rgba(255, 255, 255, .8);--weui-GLYPH-WHITE-1: rgba(255, 255, 255, .5);--weui-GLYPH-WHITE-2: rgba(255, 255, 255, .3);--weui-GLYPH-WHITE-3: #FFFFFF;--weui-GREEN-100: #74A800;--weui-GREEN-120: #5C8600;--weui-GREEN-170: #233200;--weui-GREEN-80: #8FB933;--weui-GREEN-90: #82B01A;--weui-GREEN-BG-100: #789833;--weui-GREEN-BG-110: #6B882D;--weui-GREEN-BG-130: #65802B;--weui-GREEN-BG-90: #85A247;--weui-INDIGO-100: #1196FF;--weui-INDIGO-120: #0D78CC;--weui-INDIGO-170: #052D4D;--weui-INDIGO-80: #40ABFF;--weui-INDIGO-90: #28A0FF;--weui-INDIGO-BG-100: #0D78CC;--weui-INDIGO-BG-110: #0B6BB7;--weui-INDIGO-BG-130: #09548F;--weui-INDIGO-BG-90: #2585D1;--weui-LIGHTGREEN-100: #3EB575;--weui-LIGHTGREEN-120: #31905D;--weui-LIGHTGREEN-170: #123522;--weui-LIGHTGREEN-80: #64C390;--weui-LIGHTGREEN-90: #51BC83;--weui-LIGHTGREEN-BG-100: #31905D;--weui-LIGHTGREEN-BG-110: #2C8153;--weui-LIGHTGREEN-BG-130: #226541;--weui-LIGHTGREEN-BG-90: #31905D;--weui-LINK-100: #7D90A9;--weui-LINK-120: #647387;--weui-LINK-170: #252A32;--weui-LINK-80: #97A6BA;--weui-LINK-90: #899AB1;--weui-LINKFINDER-100: #DEE9FF;--weui-MATERIAL-ATTACHMENTCOLUMN: rgba(32, 32, 32, .93);--weui-MATERIAL-NAVIGATIONBAR: rgba(18, 18, 18, .9);--weui-MATERIAL-REGULAR: rgba(37, 37, 37, .6);--weui-MATERIAL-THICK: rgba(34, 34, 34, .9);--weui-MATERIAL-THIN: rgba(95, 95, 95, .4);--weui-MATERIAL-TOOLBAR: rgba(35, 35, 35, .93);--weui-ORANGE-100: #C87D2F;--weui-ORANGE-120: #A06425;--weui-ORANGE-170: #3B250E;--weui-ORANGE-80: #D39758;--weui-ORANGE-90: #CD8943;--weui-ORANGE-BG-100: #BB6000;--weui-ORANGE-BG-110: #A85600;--weui-ORANGE-BG-130: #824300;--weui-ORANGE-BG-90: #C1701A;--weui-ORANGERED-100: #FF6146;--weui-OVERLAY: rgba(0, 0, 0, .8);--weui-OVERLAY-WHITE: rgba(242, 242, 242, .8);--weui-PURPLE-100: #8183FF;--weui-PURPLE-120: #6768CC;--weui-PURPLE-170: #26274C;--weui-PURPLE-80: #9A9BFF;--weui-PURPLE-90: #8D8FFF;--weui-PURPLE-BG-100: #6768CC;--weui-PURPLE-BG-110: #5C5DB7;--weui-PURPLE-BG-130: #48498F;--weui-PURPLE-BG-90: #7677D1;--weui-RED-100: #FA5151;--weui-RED-120: #C84040;--weui-RED-170: #4B1818;--weui-RED-80: #FB7373;--weui-RED-90: #FA6262;--weui-RED-BG-100: #CF5148;--weui-RED-BG-110: #BA4940;--weui-RED-BG-130: #913832;--weui-RED-BG-90: #D3625A;--weui-SECONDARY-BG: rgba(255, 255, 255, .1);--weui-SEPARATOR-0: rgba(255, 255, 255, .05);--weui-SEPARATOR-1: rgba(255, 255, 255, .15);--weui-STATELAYER-HOVERED: rgba(0, 0, 0, .02);--weui-STATELAYER-PRESSED: rgba(255, 255, 255, .1);--weui-STATELAYER-PRESSEDSTRENGTHENED: rgba(255, 255, 255, .2);--weui-YELLOW-100: #CC9C00;--weui-YELLOW-120: #A37C00;--weui-YELLOW-170: #3D2F00;--weui-YELLOW-80: #D6AF33;--weui-YELLOW-90: #D1A519;--weui-YELLOW-BG-100: #BF9100;--weui-YELLOW-BG-110: #AB8200;--weui-YELLOW-BG-130: #866500;--weui-YELLOW-BG-90: #C59C1A;--weui-FG-HALF: rgba(255, 255, 255, .6);--weui-RED: #FA5151;--weui-ORANGERED: #FF6146;--weui-ORANGE: #C87D2F;--weui-YELLOW: #CC9C00;--weui-GREEN: #74A800;--weui-LIGHTGREEN: #3EB575;--weui-TEXTGREEN: #259C5C;--weui-BRAND: #07C160;--weui-BLUE: #10AEFF;--weui-INDIGO: #1196FF;--weui-PURPLE: #8183FF;--weui-LINK: #7D90A9;--weui-REDORANGE: #FF6146;--weui-TAG-TEXT-BLACK: rgba(255, 255, 255, .5);--weui-TAG-BACKGROUND-BLACK: rgba(255, 255, 255, .05);--weui-WHITE: rgba(255, 255, 255, .8);--weui-FG: #fff;--weui-BG: #000;--weui-FG-5: rgba(255, 255, 255, .1);--weui-TAG-BACKGROUND-ORANGE: rgba(250, 157, 59, .1);--weui-TAG-BACKGROUND-GREEN: rgba(6, 174, 86, .1);--weui-TAG-TEXT-RED: rgba(250, 81, 81, .6);--weui-TAG-BACKGROUND-RED: rgba(250, 81, 81, .1);--weui-TAG-BACKGROUND-BLUE: rgba(16, 174, 255, .1);--weui-TAG-TEXT-ORANGE: rgba(250, 157, 59, .6);--weui-TAG-TEXT-GREEN: rgba(6, 174, 86, .6);--weui-TAG-TEXT-BLUE: rgba(16, 174, 255, .6)}}@media screen and (min-width:1024px) and (prefers-color-scheme:dark){body:not(.pages_skin_pc) .wx-root:not([data-weui-theme=light]),body:not(.pages_skin_pc) body:not([data-weui-theme=light]){--appmsgExtra-BG: #121212}}@media(prefers-color-scheme:dark){body:not([data-weui-theme=light]).my_comment_empty_data{background-color:#111}}body{--weui-elpsLine: 2;--weui-elpsFontSize: 1rem;--weui-elpsColor: rgba(0, 0, 0, .9)}body{--weui-mask-elpsLine: 2;--weui-mask-elpsLineHeight: 1.4;--weui-mask-elpsFontSize: 1rem;--weui-mask-elpsColor: rgba(0, 0, 0, .9);--weui-mask-gradient: linear-gradient(to right, rgba(255, 255, 255, 0), #ffffff 40%)}body,.wx-root{--weui-BG-6: rgba(0, 0, 0, .05);--weui-ACTIVE-MASK: rgba(0, 0, 0, .05)}body,.wx-root{--weui-BG-0: #ededed;--weui-BG-1: #f7f7f7;--weui-BG-2: #fff;--weui-BG-3: #f7f7f7;--weui-BG-4: #4c4c4c;--weui-BG-5: #fff;--weui-RED: #fa5151;--weui-ORANGERED: #ff6146;--weui-ORANGE: #fa9d3b;--weui-YELLOW: #ffc300;--weui-GREEN: #91d300;--weui-LIGHTGREEN: #95ec69;--weui-BRAND: #07c160;--weui-BLUE: #10aeff;--weui-INDIGO: #1485ee;--weui-PURPLE: #6467f0;--weui-WHITE: #fff;--weui-LINK: #576b95;--weui-TEXTGREEN: #06ae56;--weui-BG: #fff;--weui-TAG-TEXT-ORANGE: #fa9d3b;--weui-TAG-TEXT-GREEN: #06ae56;--weui-TAG-TEXT-BLUE: #10aeff;--weui-REDORANGE: #ff6146;--weui-BG-0: #EDEDED;--weui-BG-1: #F7F7F7;--weui-BG-2: #FFFFFF;--weui-BG-3: #F7F7F7;--weui-BG-4: #4C4C4C;--weui-BG-5: #FFFFFF;--weui-BLUE-100: #10AEFF;--weui-BLUE-120: #3FBEFF;--weui-BLUE-170: #B7E6FF;--weui-BLUE-80: #0C8BCC;--weui-BLUE-90: #0E9CE6;--weui-BLUE-BG-100: #48A6E2;--weui-BLUE-BG-110: #5AAFE4;--weui-BLUE-BG-130: #7FC0EA;--weui-BLUE-BG-90: #4095CB;--weui-BRAND-100: #07C160;--weui-BRAND-120: #38CD7F;--weui-BRAND-170: #B4ECCE;--weui-BRAND-80: #059A4C;--weui-BRAND-90: #06AE56;--weui-BRAND-BG-100: #2AAE67;--weui-BRAND-BG-110: #3EB575;--weui-BRAND-BG-130: #69C694;--weui-BRAND-BG-90: #259C5C;--weui-FG-0: rgba(0, 0, 0, .9);--weui-FG-0_5: rgba(0, 0, 0, .9);--weui-FG-1: rgba(0, 0, 0, .55);--weui-FG-2: rgba(0, 0, 0, .3);--weui-FG-3: rgba(0, 0, 0, .1);--weui-FG-4: rgba(0, 0, 0, .15);--weui-GLYPH-0: rgba(0, 0, 0, .9);--weui-GLYPH-1: rgba(0, 0, 0, .55);--weui-GLYPH-2: rgba(0, 0, 0, .3);--weui-GLYPH-WHITE-0: rgba(255, 255, 255, .8);--weui-GLYPH-WHITE-1: rgba(255, 255, 255, .5);--weui-GLYPH-WHITE-2: rgba(255, 255, 255, .3);--weui-GLYPH-WHITE-3: #FFFFFF;--weui-GREEN-100: #91D300;--weui-GREEN-120: #A7DB33;--weui-GREEN-170: #DEF1B3;--weui-GREEN-80: #74A800;--weui-GREEN-90: #82BD00;--weui-GREEN-BG-100: #96BE40;--weui-GREEN-BG-110: #A0C452;--weui-GREEN-BG-130: #B5D179;--weui-GREEN-BG-90: #86AA39;--weui-INDIGO-100: #1485EE;--weui-INDIGO-120: #439DF1;--weui-INDIGO-170: #B8DAF9;--weui-INDIGO-80: #106ABE;--weui-INDIGO-90: #1277D6;--weui-INDIGO-BG-100: #2B77BF;--weui-INDIGO-BG-110: #3F84C5;--weui-INDIGO-BG-130: #6BA0D2;--weui-INDIGO-BG-90: #266AAB;--weui-LIGHTGREEN-100: #95EC69;--weui-LIGHTGREEN-120: #AAEF87;--weui-LIGHTGREEN-170: #DEF9D1;--weui-LIGHTGREEN-80: #77BC54;--weui-LIGHTGREEN-90: #85D35E;--weui-LIGHTGREEN-BG-100: #72CF60;--weui-LIGHTGREEN-BG-110: #80D370;--weui-LIGHTGREEN-BG-130: #9CDD90;--weui-LIGHTGREEN-BG-90: #66B956;--weui-LINK-100: #576B95;--weui-LINK-120: #7888AA;--weui-LINK-170: #CCD2DE;--weui-LINK-80: #455577;--weui-LINK-90: #4E6085;--weui-LINKFINDER-100: #002666;--weui-MATERIAL-ATTACHMENTCOLUMN: rgba(245, 245, 245, .95);--weui-MATERIAL-NAVIGATIONBAR: rgba(237, 237, 237, .94);--weui-MATERIAL-REGULAR: rgba(247, 247, 247, .3);--weui-MATERIAL-THICK: rgba(247, 247, 247, .8);--weui-MATERIAL-THIN: rgba(255, 255, 255, .2);--weui-MATERIAL-TOOLBAR: rgba(246, 246, 246, .82);--weui-ORANGE-100: #FA9D3B;--weui-ORANGE-120: #FBB062;--weui-ORANGE-170: #FDE1C3;--weui-ORANGE-80: #C87D2F;--weui-ORANGE-90: #E08C34;--weui-ORANGE-BG-100: #EA7800;--weui-ORANGE-BG-110: #EC8519;--weui-ORANGE-BG-130: #F0A04D;--weui-ORANGE-BG-90: #D26B00;--weui-ORANGERED-100: #FF6146;--weui-OVERLAY: rgba(0, 0, 0, .5);--weui-OVERLAY-WHITE: rgba(242, 242, 242, .8);--weui-PURPLE-100: #6467F0;--weui-PURPLE-120: #8385F3;--weui-PURPLE-170: #D0D1FA;--weui-PURPLE-80: #5052C0;--weui-PURPLE-90: #595CD7;--weui-PURPLE-BG-100: #6769BA;--weui-PURPLE-BG-110: #7678C1;--weui-PURPLE-BG-130: #9496CE;--weui-PURPLE-BG-90: #5C5EA7;--weui-RED-100: #FA5151;--weui-RED-120: #FB7373;--weui-RED-170: #FDCACA;--weui-RED-80: #C84040;--weui-RED-90: #E14949;--weui-RED-BG-100: #CF5148;--weui-RED-BG-110: #D3625A;--weui-RED-BG-130: #DD847E;--weui-RED-BG-90: #B94840;--weui-SECONDARY-BG: rgba(0, 0, 0, .05);--weui-SEPARATOR-0: rgba(0, 0, 0, .1);--weui-SEPARATOR-1: rgba(0, 0, 0, .15);--weui-STATELAYER-HOVERED: rgba(0, 0, 0, .02);--weui-STATELAYER-PRESSED: rgba(0, 0, 0, .1);--weui-STATELAYER-PRESSEDSTRENGTHENED: rgba(0, 0, 0, .2);--weui-YELLOW-100: #FFC300;--weui-YELLOW-120: #FFCF33;--weui-YELLOW-170: #FFECB2;--weui-YELLOW-80: #CC9C00;--weui-YELLOW-90: #E6AF00;--weui-YELLOW-BG-100: #EFB600;--weui-YELLOW-BG-110: #F0BD19;--weui-YELLOW-BG-130: #F3CC4D;--weui-YELLOW-BG-90: #D7A400;--weui-FG-HALF: rgba(0, 0, 0, .9);--weui-RED: #FA5151;--weui-ORANGERED: #FF6146;--weui-ORANGE: #FA9D3B;--weui-YELLOW: #FFC300;--weui-GREEN: #91D300;--weui-LIGHTGREEN: #95EC69;--weui-TEXTGREEN: #06AE56;--weui-BRAND: #07C160;--weui-BLUE: #10AEFF;--weui-INDIGO: #1485EE;--weui-PURPLE: #6467F0;--weui-LINK: #576B95;--weui-TAG-TEXT-ORANGE: #FA9D3B;--weui-TAG-TEXT-GREEN: #06AE56;--weui-TAG-TEXT-BLUE: #10AEFF;--weui-REDORANGE: #FF6146;--weui-TAG-TEXT-BLACK: rgba(0, 0, 0, .5);--weui-TAG-BACKGROUND-BLACK: rgba(0, 0, 0, .05);--weui-WHITE: #FFFFFF;--weui-BG: #FFFFFF;--weui-FG: #000;--weui-FG-5: rgba(0, 0, 0, .05);--weui-TAG-BACKGROUND-ORANGE: rgba(250, 157, 59, .1);--weui-TAG-BACKGROUND-GREEN: rgba(6, 174, 86, .1);--weui-TAG-TEXT-RED: rgba(250, 81, 81, .6);--weui-TAG-BACKGROUND-RED: rgba(250, 81, 81, .1);--weui-TAG-BACKGROUND-BLUE: rgba(16, 174, 255, .1)}body,.wx-root{--weui-BG-0: #EDEDED;--weui-BG-1: #F7F7F7;--weui-BG-2: #FFFFFF;--weui-BG-3: #F7F7F7;--weui-BG-4: #4C4C4C;--weui-BG-5: #FFFFFF;--weui-BLUE-100: #10AEFF;--weui-BLUE-120: #3FBEFF;--weui-BLUE-170: #B7E6FF;--weui-BLUE-80: #0C8BCC;--weui-BLUE-90: #0E9CE6;--weui-BLUE-BG-100: #48A6E2;--weui-BLUE-BG-110: #5AAFE4;--weui-BLUE-BG-130: #7FC0EA;--weui-BLUE-BG-90: #4095CB;--weui-BRAND-100: #07C160;--weui-BRAND-120: #38CD7F;--weui-BRAND-170: #B4ECCE;--weui-BRAND-80: #059A4C;--weui-BRAND-90: #06AE56;--weui-BRAND-BG-100: #2AAE67;--weui-BRAND-BG-110: #3EB575;--weui-BRAND-BG-130: #69C694;--weui-BRAND-BG-90: #259C5C;--weui-FG-0: rgba(0, 0, 0, .9);--weui-FG-0_5: rgba(0, 0, 0, .9);--weui-FG-1: rgba(0, 0, 0, .55);--weui-FG-2: rgba(0, 0, 0, .3);--weui-FG-3: rgba(0, 0, 0, .1);--weui-FG-4: rgba(0, 0, 0, .15);--weui-GLYPH-0: rgba(0, 0, 0, .9);--weui-GLYPH-1: rgba(0, 0, 0, .55);--weui-GLYPH-2: rgba(0, 0, 0, .3);--weui-GLYPH-WHITE-0: rgba(255, 255, 255, .8);--weui-GLYPH-WHITE-1: rgba(255, 255, 255, .5);--weui-GLYPH-WHITE-2: rgba(255, 255, 255, .3);--weui-GLYPH-WHITE-3: #FFFFFF;--weui-GREEN-100: #91D300;--weui-GREEN-120: #A7DB33;--weui-GREEN-170: #DEF1B3;--weui-GREEN-80: #74A800;--weui-GREEN-90: #82BD00;--weui-GREEN-BG-100: #96BE40;--weui-GREEN-BG-110: #A0C452;--weui-GREEN-BG-130: #B5D179;--weui-GREEN-BG-90: #86AA39;--weui-INDIGO-100: #1485EE;--weui-INDIGO-120: #439DF1;--weui-INDIGO-170: #B8DAF9;--weui-INDIGO-80: #106ABE;--weui-INDIGO-90: #1277D6;--weui-INDIGO-BG-100: #2B77BF;--weui-INDIGO-BG-110: #3F84C5;--weui-INDIGO-BG-130: #6BA0D2;--weui-INDIGO-BG-90: #266AAB;--weui-LIGHTGREEN-100: #95EC69;--weui-LIGHTGREEN-120: #AAEF87;--weui-LIGHTGREEN-170: #DEF9D1;--weui-LIGHTGREEN-80: #77BC54;--weui-LIGHTGREEN-90: #85D35E;--weui-LIGHTGREEN-BG-100: #72CF60;--weui-LIGHTGREEN-BG-110: #80D370;--weui-LIGHTGREEN-BG-130: #9CDD90;--weui-LIGHTGREEN-BG-90: #66B956;--weui-LINK-100: #576B95;--weui-LINK-120: #7888AA;--weui-LINK-170: #CCD2DE;--weui-LINK-80: #455577;--weui-LINK-90: #4E6085;--weui-LINKFINDER-100: #002666;--weui-MATERIAL-ATTACHMENTCOLUMN: rgba(245, 245, 245, .95);--weui-MATERIAL-NAVIGATIONBAR: rgba(237, 237, 237, .94);--weui-MATERIAL-REGULAR: rgba(247, 247, 247, .3);--weui-MATERIAL-THICK: rgba(247, 247, 247, .8);--weui-MATERIAL-THIN: rgba(255, 255, 255, .2);--weui-MATERIAL-TOOLBAR: rgba(246, 246, 246, .82);--weui-ORANGE-100: #FA9D3B;--weui-ORANGE-120: #FBB062;--weui-ORANGE-170: #FDE1C3;--weui-ORANGE-80: #C87D2F;--weui-ORANGE-90: #E08C34;--weui-ORANGE-BG-100: #EA7800;--weui-ORANGE-BG-110: #EC8519;--weui-ORANGE-BG-130: #F0A04D;--weui-ORANGE-BG-90: #D26B00;--weui-ORANGERED-100: #FF6146;--weui-OVERLAY: rgba(0, 0, 0, .5);--weui-OVERLAY-WHITE: rgba(242, 242, 242, .8);--weui-PURPLE-100: #6467F0;--weui-PURPLE-120: #8385F3;--weui-PURPLE-170: #D0D1FA;--weui-PURPLE-80: #5052C0;--weui-PURPLE-90: #595CD7;--weui-PURPLE-BG-100: #6769BA;--weui-PURPLE-BG-110: #7678C1;--weui-PURPLE-BG-130: #9496CE;--weui-PURPLE-BG-90: #5C5EA7;--weui-RED-100: #FA5151;--weui-RED-120: #FB7373;--weui-RED-170: #FDCACA;--weui-RED-80: #C84040;--weui-RED-90: #E14949;--weui-RED-BG-100: #CF5148;--weui-RED-BG-110: #D3625A;--weui-RED-BG-130: #DD847E;--weui-RED-BG-90: #B94840;--weui-SECONDARY-BG: rgba(0, 0, 0, .05);--weui-SEPARATOR-0: rgba(0, 0, 0, .1);--weui-SEPARATOR-1: rgba(0, 0, 0, .15);--weui-STATELAYER-HOVERED: rgba(0, 0, 0, .02);--weui-STATELAYER-PRESSED: rgba(0, 0, 0, .1);--weui-STATELAYER-PRESSEDSTRENGTHENED: rgba(0, 0, 0, .2);--weui-YELLOW-100: #FFC300;--weui-YELLOW-120: #FFCF33;--weui-YELLOW-170: #FFECB2;--weui-YELLOW-80: #CC9C00;--weui-YELLOW-90: #E6AF00;--weui-YELLOW-BG-100: #EFB600;--weui-YELLOW-BG-110: #F0BD19;--weui-YELLOW-BG-130: #F3CC4D;--weui-YELLOW-BG-90: #D7A400;--weui-FG-HALF: rgba(0, 0, 0, .9);--weui-RED: #FA5151;--weui-ORANGERED: #FF6146;--weui-ORANGE: #FA9D3B;--weui-YELLOW: #FFC300;--weui-GREEN: #91D300;--weui-LIGHTGREEN: #95EC69;--weui-TEXTGREEN: #06AE56;--weui-BRAND: #07C160;--weui-BLUE: #10AEFF;--weui-INDIGO: #1485EE;--weui-PURPLE: #6467F0;--weui-LINK: #576B95;--weui-TAG-TEXT-ORANGE: #FA9D3B;--weui-TAG-TEXT-GREEN: #06AE56;--weui-TAG-TEXT-BLUE: #10AEFF;--weui-REDORANGE: #FF6146;--weui-TAG-TEXT-BLACK: rgba(0, 0, 0, .5);--weui-TAG-BACKGROUND-BLACK: rgba(0, 0, 0, .05);--weui-WHITE: #FFFFFF;--weui-BG: #FFFFFF;--weui-FG: #000;--weui-FG-5: rgba(0, 0, 0, .05);--weui-TAG-BACKGROUND-ORANGE: rgba(250, 157, 59, .1);--weui-TAG-BACKGROUND-GREEN: rgba(6, 174, 86, .1);--weui-TAG-TEXT-RED: rgba(250, 81, 81, .6);--weui-TAG-BACKGROUND-RED: rgba(250, 81, 81, .1);--weui-TAG-BACKGROUND-BLUE: rgba(16, 174, 255, .1)}.wx-root,body{--weui-BTN-ACTIVE-MASK: rgba(0, 0, 0, .1)}.wx-root,body{--weui-BTN-DEFAULT-ACTIVE-BG: #e6e6e6}.wx-root,body{--weui-DIALOG-LINE-COLOR: rgba(0, 0, 0, .1)}.wx-root,body{--weui-BG-COLOR-ACTIVE: #ececec}@media(prefers-color-scheme:dark){.wx-root:not([data-weui-theme=light]),body:not([data-weui-theme=light]){--weui-BTN-ACTIVE-MASK: rgba(255, 255, 255, .1)}}@media(prefers-color-scheme:dark){.wx-root:not([data-weui-theme=light]),body:not([data-weui-theme=light]){--weui-BTN-DEFAULT-ACTIVE-BG: rgba(255, 255, 255, .126)}}@media(prefers-color-scheme:dark){.wx-root:not([data-weui-theme=light]),body:not([data-weui-theme=light]){--weui-DIALOG-LINE-COLOR: rgba(255, 255, 255, .1)}}@media(prefers-color-scheme:dark){.wx-root:not([data-weui-theme=light]),body:not([data-weui-theme=light]){--APPMSGCARD-BG: #1E1E1E}}@media(prefers-color-scheme:dark){.wx-root:not([data-weui-theme=light]),body:not([data-weui-theme=light]){--APPMSGCARD-LINE-BG: rgba(255, 255, 255, .07)}}@media(prefers-color-scheme:dark){.wx-root:not([data-weui-theme=light]),body:not([data-weui-theme=light]){--weui-BG-COLOR-ACTIVE: #373737}}@media(prefers-color-scheme:dark){.wx-root:not([data-weui-theme=light]),body:not([data-weui-theme=light]){--weui-BG-6: rgba(255, 255, 255, .1);--weui-ACTIVE-MASK: rgba(255, 255, 255, .1)}}@media(prefers-color-scheme:dark){.wx-root:not([data-weui-theme=light]),body:not([data-weui-theme=light]){--weui-BG-0: #111;--weui-BG-1: #1e1e1e;--weui-BG-5: #2c2c2c;--weui-RED: #fa5151;--weui-ORANGERED: #ff6146;--weui-ORANGE: #c87d2f;--weui-YELLOW: #cc9c00;--weui-GREEN: #74a800;--weui-LIGHTGREEN: #3eb575;--weui-BRAND: #07c160;--weui-BLUE: #10aeff;--weui-INDIGO: #1196ff;--weui-PURPLE: #8183ff;--weui-LINK: #7d90a9;--weui-TEXTGREEN: #259c5c;--weui-REDORANGE: #ff6146;--weui-BG-0: #111111;--weui-BG-1: #1E1E1E;--weui-BG-2: #191919;--weui-BG-3: #202020;--weui-BG-4: #404040;--weui-BG-5: #2C2C2C;--weui-BLUE-100: #10AEFF;--weui-BLUE-120: #0C8BCC;--weui-BLUE-170: #04344D;--weui-BLUE-80: #3FBEFF;--weui-BLUE-90: #28B6FF;--weui-BLUE-BG-100: #48A6E2;--weui-BLUE-BG-110: #4095CB;--weui-BLUE-BG-130: #32749E;--weui-BLUE-BG-90: #5AAFE4;--weui-BRAND-100: #07C160;--weui-BRAND-120: #059A4C;--weui-BRAND-170: #023A1C;--weui-BRAND-80: #38CD7F;--weui-BRAND-90: #20C770;--weui-BRAND-BG-100: #2AAE67;--weui-BRAND-BG-110: #259C5C;--weui-BRAND-BG-130: #1D7A48;--weui-BRAND-BG-90: #3EB575;--weui-FG-0: rgba(255, 255, 255, .8);--weui-FG-0_5: rgba(255, 255, 255, .6);--weui-FG-1: rgba(255, 255, 255, .5);--weui-FG-2: rgba(255, 255, 255, .3);--weui-FG-3: rgba(255, 255, 255, .1);--weui-FG-4: rgba(255, 255, 255, .15);--weui-GLYPH-0: rgba(255, 255, 255, .8);--weui-GLYPH-1: rgba(255, 255, 255, .5);--weui-GLYPH-2: rgba(255, 255, 255, .3);--weui-GLYPH-WHITE-0: rgba(255, 255, 255, .8);--weui-GLYPH-WHITE-1: rgba(255, 255, 255, .5);--weui-GLYPH-WHITE-2: rgba(255, 255, 255, .3);--weui-GLYPH-WHITE-3: #FFFFFF;--weui-GREEN-100: #74A800;--weui-GREEN-120: #5C8600;--weui-GREEN-170: #233200;--weui-GREEN-80: #8FB933;--weui-GREEN-90: #82B01A;--weui-GREEN-BG-100: #789833;--weui-GREEN-BG-110: #6B882D;--weui-GREEN-BG-130: #65802B;--weui-GREEN-BG-90: #85A247;--weui-INDIGO-100: #1196FF;--weui-INDIGO-120: #0D78CC;--weui-INDIGO-170: #052D4D;--weui-INDIGO-80: #40ABFF;--weui-INDIGO-90: #28A0FF;--weui-INDIGO-BG-100: #0D78CC;--weui-INDIGO-BG-110: #0B6BB7;--weui-INDIGO-BG-130: #09548F;--weui-INDIGO-BG-90: #2585D1;--weui-LIGHTGREEN-100: #3EB575;--weui-LIGHTGREEN-120: #31905D;--weui-LIGHTGREEN-170: #123522;--weui-LIGHTGREEN-80: #64C390;--weui-LIGHTGREEN-90: #51BC83;--weui-LIGHTGREEN-BG-100: #31905D;--weui-LIGHTGREEN-BG-110: #2C8153;--weui-LIGHTGREEN-BG-130: #226541;--weui-LIGHTGREEN-BG-90: #31905D;--weui-LINK-100: #7D90A9;--weui-LINK-120: #647387;--weui-LINK-170: #252A32;--weui-LINK-80: #97A6BA;--weui-LINK-90: #899AB1;--weui-LINKFINDER-100: #DEE9FF;--weui-MATERIAL-ATTACHMENTCOLUMN: rgba(32, 32, 32, .93);--weui-MATERIAL-NAVIGATIONBAR: rgba(18, 18, 18, .9);--weui-MATERIAL-REGULAR: rgba(37, 37, 37, .6);--weui-MATERIAL-THICK: rgba(34, 34, 34, .9);--weui-MATERIAL-THIN: rgba(95, 95, 95, .4);--weui-MATERIAL-TOOLBAR: rgba(35, 35, 35, .93);--weui-ORANGE-100: #C87D2F;--weui-ORANGE-120: #A06425;--weui-ORANGE-170: #3B250E;--weui-ORANGE-80: #D39758;--weui-ORANGE-90: #CD8943;--weui-ORANGE-BG-100: #BB6000;--weui-ORANGE-BG-110: #A85600;--weui-ORANGE-BG-130: #824300;--weui-ORANGE-BG-90: #C1701A;--weui-ORANGERED-100: #FF6146;--weui-OVERLAY: rgba(0, 0, 0, .8);--weui-OVERLAY-WHITE: rgba(242, 242, 242, .8);--weui-PURPLE-100: #8183FF;--weui-PURPLE-120: #6768CC;--weui-PURPLE-170: #26274C;--weui-PURPLE-80: #9A9BFF;--weui-PURPLE-90: #8D8FFF;--weui-PURPLE-BG-100: #6768CC;--weui-PURPLE-BG-110: #5C5DB7;--weui-PURPLE-BG-130: #48498F;--weui-PURPLE-BG-90: #7677D1;--weui-RED-100: #FA5151;--weui-RED-120: #C84040;--weui-RED-170: #4B1818;--weui-RED-80: #FB7373;--weui-RED-90: #FA6262;--weui-RED-BG-100: #CF5148;--weui-RED-BG-110: #BA4940;--weui-RED-BG-130: #913832;--weui-RED-BG-90: #D3625A;--weui-SECONDARY-BG: rgba(255, 255, 255, .1);--weui-SEPARATOR-0: rgba(255, 255, 255, .05);--weui-SEPARATOR-1: rgba(255, 255, 255, .15);--weui-STATELAYER-HOVERED: rgba(0, 0, 0, .02);--weui-STATELAYER-PRESSED: rgba(255, 255, 255, .1);--weui-STATELAYER-PRESSEDSTRENGTHENED: rgba(255, 255, 255, .2);--weui-YELLOW-100: #CC9C00;--weui-YELLOW-120: #A37C00;--weui-YELLOW-170: #3D2F00;--weui-YELLOW-80: #D6AF33;--weui-YELLOW-90: #D1A519;--weui-YELLOW-BG-100: #BF9100;--weui-YELLOW-BG-110: #AB8200;--weui-YELLOW-BG-130: #866500;--weui-YELLOW-BG-90: #C59C1A;--weui-FG-HALF: rgba(255, 255, 255, .6);--weui-RED: #FA5151;--weui-ORANGERED: #FF6146;--weui-ORANGE: #C87D2F;--weui-YELLOW: #CC9C00;--weui-GREEN: #74A800;--weui-LIGHTGREEN: #3EB575;--weui-TEXTGREEN: #259C5C;--weui-BRAND: #07C160;--weui-BLUE: #10AEFF;--weui-INDIGO: #1196FF;--weui-PURPLE: #8183FF;--weui-LINK: #7D90A9;--weui-REDORANGE: #FF6146;--weui-TAG-TEXT-BLACK: rgba(255, 255, 255, .5);--weui-TAG-BACKGROUND-BLACK: rgba(255, 255, 255, .05);--weui-WHITE: rgba(255, 255, 255, .8);--weui-FG: #fff;--weui-BG: #000;--weui-FG-5: rgba(255, 255, 255, .1);--weui-TAG-BACKGROUND-ORANGE: rgba(250, 157, 59, .1);--weui-TAG-BACKGROUND-GREEN: rgba(6, 174, 86, .1);--weui-TAG-TEXT-RED: rgba(250, 81, 81, .6);--weui-TAG-BACKGROUND-RED: rgba(250, 81, 81, .1);--weui-TAG-BACKGROUND-BLUE: rgba(16, 174, 255, .1);--weui-TAG-TEXT-ORANGE: rgba(250, 157, 59, .6);--weui-TAG-TEXT-GREEN: rgba(6, 174, 86, .6);--weui-TAG-TEXT-BLUE: rgba(16, 174, 255, .6)}}@media(prefers-color-scheme:dark){.wx-root:not([data-weui-theme=light]),body:not([data-weui-theme=light]){--discussInput-BG: rgba(255, 255, 255, .03)}}@media(prefers-color-scheme:dark){.wx-root:not([data-weui-theme=light]),body:not([data-weui-theme=light]){--nickName-FG: #959595}}@media(prefers-color-scheme:dark){.wx-root:not([data-weui-theme=light]),body:not([data-weui-theme=light]){--weui-BG-0: #111111;--weui-BG-1: #1E1E1E;--weui-BG-2: #191919;--weui-BG-3: #202020;--weui-BG-4: #404040;--weui-BG-5: #2C2C2C;--weui-BLUE-100: #10AEFF;--weui-BLUE-120: #0C8BCC;--weui-BLUE-170: #04344D;--weui-BLUE-80: #3FBEFF;--weui-BLUE-90: #28B6FF;--weui-BLUE-BG-100: #48A6E2;--weui-BLUE-BG-110: #4095CB;--weui-BLUE-BG-130: #32749E;--weui-BLUE-BG-90: #5AAFE4;--weui-BRAND-100: #07C160;--weui-BRAND-120: #059A4C;--weui-BRAND-170: #023A1C;--weui-BRAND-80: #38CD7F;--weui-BRAND-90: #20C770;--weui-BRAND-BG-100: #2AAE67;--weui-BRAND-BG-110: #259C5C;--weui-BRAND-BG-130: #1D7A48;--weui-BRAND-BG-90: #3EB575;--weui-FG-0: rgba(255, 255, 255, .8);--weui-FG-0_5: rgba(255, 255, 255, .6);--weui-FG-1: rgba(255, 255, 255, .5);--weui-FG-2: rgba(255, 255, 255, .3);--weui-FG-3: rgba(255, 255, 255, .1);--weui-FG-4: rgba(255, 255, 255, .15);--weui-GLYPH-0: rgba(255, 255, 255, .8);--weui-GLYPH-1: rgba(255, 255, 255, .5);--weui-GLYPH-2: rgba(255, 255, 255, .3);--weui-GLYPH-WHITE-0: rgba(255, 255, 255, .8);--weui-GLYPH-WHITE-1: rgba(255, 255, 255, .5);--weui-GLYPH-WHITE-2: rgba(255, 255, 255, .3);--weui-GLYPH-WHITE-3: #FFFFFF;--weui-GREEN-100: #74A800;--weui-GREEN-120: #5C8600;--weui-GREEN-170: #233200;--weui-GREEN-80: #8FB933;--weui-GREEN-90: #82B01A;--weui-GREEN-BG-100: #789833;--weui-GREEN-BG-110: #6B882D;--weui-GREEN-BG-130: #65802B;--weui-GREEN-BG-90: #85A247;--weui-INDIGO-100: #1196FF;--weui-INDIGO-120: #0D78CC;--weui-INDIGO-170: #052D4D;--weui-INDIGO-80: #40ABFF;--weui-INDIGO-90: #28A0FF;--weui-INDIGO-BG-100: #0D78CC;--weui-INDIGO-BG-110: #0B6BB7;--weui-INDIGO-BG-130: #09548F;--weui-INDIGO-BG-90: #2585D1;--weui-LIGHTGREEN-100: #3EB575;--weui-LIGHTGREEN-120: #31905D;--weui-LIGHTGREEN-170: #123522;--weui-LIGHTGREEN-80: #64C390;--weui-LIGHTGREEN-90: #51BC83;--weui-LIGHTGREEN-BG-100: #31905D;--weui-LIGHTGREEN-BG-110: #2C8153;--weui-LIGHTGREEN-BG-130: #226541;--weui-LIGHTGREEN-BG-90: #31905D;--weui-LINK-100: #7D90A9;--weui-LINK-120: #647387;--weui-LINK-170: #252A32;--weui-LINK-80: #97A6BA;--weui-LINK-90: #899AB1;--weui-LINKFINDER-100: #DEE9FF;--weui-MATERIAL-ATTACHMENTCOLUMN: rgba(32, 32, 32, .93);--weui-MATERIAL-NAVIGATIONBAR: rgba(18, 18, 18, .9);--weui-MATERIAL-REGULAR: rgba(37, 37, 37, .6);--weui-MATERIAL-THICK: rgba(34, 34, 34, .9);--weui-MATERIAL-THIN: rgba(95, 95, 95, .4);--weui-MATERIAL-TOOLBAR: rgba(35, 35, 35, .93);--weui-ORANGE-100: #C87D2F;--weui-ORANGE-120: #A06425;--weui-ORANGE-170: #3B250E;--weui-ORANGE-80: #D39758;--weui-ORANGE-90: #CD8943;--weui-ORANGE-BG-100: #BB6000;--weui-ORANGE-BG-110: #A85600;--weui-ORANGE-BG-130: #824300;--weui-ORANGE-BG-90: #C1701A;--weui-ORANGERED-100: #FF6146;--weui-OVERLAY: rgba(0, 0, 0, .8);--weui-OVERLAY-WHITE: rgba(242, 242, 242, .8);--weui-PURPLE-100: #8183FF;--weui-PURPLE-120: #6768CC;--weui-PURPLE-170: #26274C;--weui-PURPLE-80: #9A9BFF;--weui-PURPLE-90: #8D8FFF;--weui-PURPLE-BG-100: #6768CC;--weui-PURPLE-BG-110: #5C5DB7;--weui-PURPLE-BG-130: #48498F;--weui-PURPLE-BG-90: #7677D1;--weui-RED-100: #FA5151;--weui-RED-120: #C84040;--weui-RED-170: #4B1818;--weui-RED-80: #FB7373;--weui-RED-90: #FA6262;--weui-RED-BG-100: #CF5148;--weui-RED-BG-110: #BA4940;--weui-RED-BG-130: #913832;--weui-RED-BG-90: #D3625A;--weui-SECONDARY-BG: rgba(255, 255, 255, .1);--weui-SEPARATOR-0: rgba(255, 255, 255, .05);--weui-SEPARATOR-1: rgba(255, 255, 255, .15);--weui-STATELAYER-HOVERED: rgba(0, 0, 0, .02);--weui-STATELAYER-PRESSED: rgba(255, 255, 255, .1);--weui-STATELAYER-PRESSEDSTRENGTHENED: rgba(255, 255, 255, .2);--weui-YELLOW-100: #CC9C00;--weui-YELLOW-120: #A37C00;--weui-YELLOW-170: #3D2F00;--weui-YELLOW-80: #D6AF33;--weui-YELLOW-90: #D1A519;--weui-YELLOW-BG-100: #BF9100;--weui-YELLOW-BG-110: #AB8200;--weui-YELLOW-BG-130: #866500;--weui-YELLOW-BG-90: #C59C1A;--weui-FG-HALF: rgba(255, 255, 255, .6);--weui-RED: #FA5151;--weui-ORANGERED: #FF6146;--weui-ORANGE: #C87D2F;--weui-YELLOW: #CC9C00;--weui-GREEN: #74A800;--weui-LIGHTGREEN: #3EB575;--weui-TEXTGREEN: #259C5C;--weui-BRAND: #07C160;--weui-BLUE: #10AEFF;--weui-INDIGO: #1196FF;--weui-PURPLE: #8183FF;--weui-LINK: #7D90A9;--weui-REDORANGE: #FF6146;--weui-TAG-TEXT-BLACK: rgba(255, 255, 255, .5);--weui-TAG-BACKGROUND-BLACK: rgba(255, 255, 255, .05);--weui-WHITE: rgba(255, 255, 255, .8);--weui-FG: #fff;--weui-BG: #000;--weui-FG-5: rgba(255, 255, 255, .1);--weui-TAG-BACKGROUND-ORANGE: rgba(250, 157, 59, .1);--weui-TAG-BACKGROUND-GREEN: rgba(6, 174, 86, .1);--weui-TAG-TEXT-RED: rgba(250, 81, 81, .6);--weui-TAG-BACKGROUND-RED: rgba(250, 81, 81, .1);--weui-TAG-BACKGROUND-BLUE: rgba(16, 174, 255, .1);--weui-TAG-TEXT-ORANGE: rgba(250, 157, 59, .6);--weui-TAG-TEXT-GREEN: rgba(6, 174, 86, .6);--weui-TAG-TEXT-BLUE: rgba(16, 174, 255, .6)}}@media(prefers-color-scheme:dark){.wx-root:not([data-weui-theme=light]),body:not([data-weui-theme=light]){--appmsgExtra-BG: #121212}}\u003c/style\u003e\u003cstyle\u003ediv.autoTypeSetting24psection \u003e p,div.autoTypeSetting24psection \u003e section{margin-bottom: 24px;}\u003c/style\u003e\u003c/div\u003e

本文来自作者[涵舒少爷]投稿,不代表蔚蓝之海立场,如若转载,请注明出处:https://foryh.cn/wiki/31317.html

(736)

文章推荐

  • 张继科丁宁现任男友 马龙张继科丁宁三角恋

    张继科丁宁现任男友张继科不是丁宁现任男友。丁宁是中国乒乓球运动员中的大满贯获得者,她在里约奥运会上夺得了女子单打金牌,还获得了世界杯、世锦赛冠军,还曾经是中国女队的队长。到了东京奥运会周期,因为她年龄偏大,身体伤病多,所以没有参加奥运会。为什么张继科丁宁不打球了张继科、丁宁因为伤病、健康、年龄偏大,

    2024年08月21日
    304
  • 2024网络直播_良心企业,值得支持_安卓版420.425

    2024网络直播随着互联网技术的飞速发展,网络直播已经成为人们日常生活中不可或缺的一部分。2024年,网络直播行业将迎来新的发展机遇和挑战。首先,5G技术的普及将为网络直播带来革命性的变化。5G网络的高速率、低延迟特性将使得直播画面更加清晰流畅,用户体验大幅提升。同时,5G网络的广泛覆盖也将使得更多

    2024年11月23日
    816
  • 2024澳门管家婆资_详细解答解释落实_iPad14.06.89

    随着金融市场的不断变化和投资者对理财的需求日益增加,如何在复杂多变的市场环境中获得可持续的财富增值,成为了众多投资者最为关心的问题。2024年,澳门管家婆资应运而生,凭借其强大的市场分析能力和精准的投资建议,迅速成为理财市场的热点。澳门管家婆资不仅继承了传统的财富管理智慧,更引入了先进的人工智能和

    2024年11月24日
    33
  • 新澳精准资料免费提供630期_作答解释落实_安卓版385.477

    在如今竞争激烈的时代,如何把握机会、赢得人生?答案或许就在于拥有准确的信息和精准的判断。对于广大彩民朋友而言,拥有一份靠谱的资料尤为关键。而这一次,我们为您免费提供新澳精准资料630期,让您轻松掌握财富密码,赢在起跑线上!什么是新澳精准资料?新澳精准资料是专为投注爱好者打造的高质量信息参考资料,

    2024年12月09日
    53
  • 澳门一肖一码100%精准王中王_精选解释落实将深度解析_V31.35.60

    澳门一肖一码100%精准王中王:开启你的财富之门在充满竞争与机会的世界里,如何从众多选择中脱颖而出,获得属于自己的财富,始终是每个人梦寐以求的目标。而在博彩与彩票领域,澳门无疑是全球最具吸引力和竞争力的市场之一。作为亚洲的博彩之都,澳门的每一个彩票选择都蕴含着巨大的财富潜力。尤其是在澳门的“一肖一

    2024年12月10日
    27
  • 2024一肖一码100%_一句引发热议_实用版306.751

    在如今快速发展的信息社会里,追求精准、快速、有效的财富增值方式已成为人们的共同需求。而提到精准预测和策略规划,“2024一肖一码100%”的概念无疑成为焦点,它既是一种象征精准的方法论,也是助力财富升级的利器。本文将为你揭秘这一方法的核心逻辑及其背后的独特魅力。一、何为2024一肖一码100%?

    2024年12月12日
    30
  • 最准一码一肖100%精准阿里云_最佳选择_V09.20.03

    在当今大数据与人工智能主导的时代,精准预测正成为个人与企业决策的核心需求。而作为全球领先的云计算服务提供商,阿里云以其强大的技术实力和广泛的应用场景,推出了最准一码一肖100%精准的解决方案,帮助用户以简单高效的方式实现智能化的突破。精准预测的时代背景无论是商业场景中的市场趋势分析,还是生活场景

    2024年12月14日
    29
  • 新澳一肖一码100免费资枓_作答解释落实_iPad39.26.47

    在现代生活中,彩票不仅是一种娱乐方式,更是一种增加收入的渠道。如何在这片数字的海洋中找到属于自己的“财富钥匙”?这正是“新澳一肖一码100免费资枓”为广大彩民提供帮助的初心。通过精准的数据分析和可靠的内部信息,我们致力于成为您在彩票领域的得力助手,助您实现财富梦想。1.什么是新澳一肖一码100免费

    2024年12月20日
    26
  • 2024新澳历史开奖记录_最新答案解释落实_手机版709.945

    在彩票领域,“新澳历史开奖记录”始终是一项备受关注的数据宝库。对于彩民来说,这不仅是过去数据的集合,更是一把打开财富之门的钥匙。2024年的新澳开奖记录正在以其独特的趋势和规律吸引着无数关注的目光。如果你希望在彩票投注中更加游刃有余,掌握历史数据无疑是至关重要的一步。新澳历史开奖记录的价值与意义

    2024年12月23日
    17
  • 澳门9点内部码群_一句引发热议_安卓版710.804

    在澳门这个充满光彩与刺激的博彩圣地,精通数字与概率是一门被推崇的艺术。对于许多热衷博彩的人来说,能够获得更高概率的“内幕信息”往往能左右他们的成败。因此,澳门9点内部码群应运而生,为博彩爱好者提供一个绝佳的互动和信息共享平台。什么是澳门9点内部码群?简而言之,这是一个由资深玩家、博彩专家和技术人员

    2024年11月29日
    31

发表回复

本站作者后才能评论

评论列表(4条)

  • 涵舒少爷
    涵舒少爷 2025年01月25日

    我是蔚蓝之海的签约作者“涵舒少爷”!

  • 涵舒少爷
    涵舒少爷 2025年01月25日

    希望本篇文章《问世两周年看ChatGPT发展史:从基础神经元到多模态智能体_腾讯新闻》能对你有所帮助!

  • 涵舒少爷
    涵舒少爷 2025年01月25日

    本站[蔚蓝之海]内容主要涵盖:生活百科,小常识等内容......

  • 涵舒少爷
    涵舒少爷 2025年01月25日

    本文概览:############### \u003cdiv class="rich_media_content autoTypeSetting24psection"\u003e\u00...

    联系我们

    邮件:蔚蓝之海@sina.com

    工作时间:周一至周五,9:30-18:30,节假日休息

    关注我们