<!DOCTYPE html><html lang="zh-CN"><head>
<meta charset="UTF-8"/>
<meta name="viewport" content="width=device-width, initial-scale=1.0"/>
<title>语言理解的"出口假设"：从神经科学到AI的深度探索</title>
<script src="https://cdn.tailwindcss.com"></script>
<link href="https://fonts.googleapis.com/css2?family=Crimson+Text:ital,wght@0,400;0,600;1,400&family=Inter:wght@300;400;500;600&display=swap" rel="stylesheet"/>
<link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.0/css/all.min.css"/>
<script src="https://cdn.jsdelivr.net/npm/mermaid/dist/mermaid.min.js"></script>
<style>
body {
font-family: 'Inter', sans-serif;
overflow-x: hidden;
}
.serif { font-family: 'Crimson Text', serif; }
html { scroll-behavior: smooth; }
/* Mermaid diagram styling */
.mermaid-container {
display: flex;
justify-content: center;
min-height: 300px;
max-height: 800px;
background: #ffffff;
border: 2px solid #e5e7eb;
border-radius: 12px;
padding: 30px;
margin: 30px 0;
box-shadow: 0 8px 25px rgba(0, 0, 0, 0.08);
position: relative;
overflow: hidden;
}
.mermaid-container .mermaid {
width: 100%;
max-width: 100%;
height: 100%;
cursor: grab;
transition: transform 0.3s ease;
transform-origin: center center;
display: flex;
justify-content: center;
align-items: center;
touch-action: none; /* 防止触摸设备上的默认行为 */
-webkit-user-select: none; /* 防止文本选择 */
-moz-user-select: none;
-ms-user-select: none;
user-select: none;
}
.mermaid-container .mermaid svg {
max-width: 100%;
height: 100%;
display: block;
margin: 0 auto;
}
.mermaid-container .mermaid:active {
cursor: grabbing;
}
.mermaid-container.zoomed .mermaid {
height: 100%;
width: 100%;
cursor: grab;
}
.mermaid-controls {
position: absolute;
top: 15px;
right: 15px;
display: flex;
gap: 10px;
z-index: 20;
background: rgba(255, 255, 255, 0.95);
padding: 8px;
border-radius: 8px;
box-shadow: 0 2px 8px rgba(0, 0, 0, 0.1);
}
.mermaid-control-btn {
background: #ffffff;
border: 1px solid #d1d5db;
border-radius: 6px;
padding: 10px;
cursor: pointer;
transition: all 0.2s ease;
color: #374151;
font-size: 14px;
min-width: 36px;
height: 36px;
text-align: center;
display: flex;
align-items: center;
justify-content: center;
}
.mermaid-control-btn:hover {
background: #f8fafc;
border-color: #3b82f6;
color: #3b82f6;
transform: translateY(-1px);
}
.mermaid-control-btn:active {
transform: scale(0.95);
}
/* Enhanced mermaid diagram text contrast and styling */
.mermaid .node rect,
.mermaid .node circle,
.mermaid .node ellipse,
.mermaid .node polygon {
stroke: #374151 !important;
stroke-width: 2px !important;
}
.mermaid .node .label {
color: #1f2937 !important;
font-weight: 600 !important;
font-size: 13px !important;
font-family: 'Inter', sans-serif !important;
}
.mermaid .edgeLabel {
background-color: rgba(255, 255, 255, 0.95) !important;
color: #374151 !important;
font-weight: 500 !important;
font-size: 12px !important;
padding: 4px 8px !important;
border-radius: 4px !important;
border: 1px solid #d1d5db !important;
font-family: 'Inter', sans-serif !important;
}
.mermaid .edge-pattern-solid {
stroke: #6b7280 !important;
stroke-width: 2px !important;
}
.mermaid .edge-pattern-dotted {
stroke: #9ca3af !important;
stroke-width: 2px !important;
}
/* Specific node color enhancements for better contrast */
.mermaid .node[style*="fill:#dbeafe"] rect,
.mermaid .node[style*="fill:#dbeafe"] circle,
.mermaid .node[style*="fill:#dbeafe"] polygon {
fill: #dbeafe !important;
stroke: #1d4ed8 !important;
stroke-width: 2px !important;
}
.mermaid .node[style*="fill:#dbeafe"] .label {
color: #1e3a8a !important;
font-weight: 600 !important;
}
.mermaid .node[style*="fill:#f3f4f6"] rect,
.mermaid .node[style*="fill:#f3f4f6"] circle,
.mermaid .node[style*="fill:#f3f4f6"] polygon {
fill: #f3f4f6 !important;
stroke: #4b5563 !important;
stroke-width: 2px !important;
}
.mermaid .node[style*="fill:#f3f4f6"] .label {
color: #1f2937 !important;
font-weight: 600 !important;
}
.mermaid .node[style*="fill:#fef3c7"] rect,
.mermaid .node[style*="fill:#fef3c7"] circle,
.mermaid .node[style*="fill:#fef3c7"] polygon {
fill: #fef3c7 !important;
stroke: #d97706 !important;
stroke-width: 2px !important;
}
.mermaid .node[style*="fill:#fef3c7"] .label {
color: #92400e !important;
font-weight: 600 !important;
}
.mermaid .node[style*="fill:#d1fae5"] rect,
.mermaid .node[style*="fill:#d1fae5"] circle,
.mermaid .node[style*="fill:#d1fae5"] polygon {
fill: #d1fae5 !important;
stroke: #059669 !important;
stroke-width: 2px !important;
}
.mermaid .node[style*="fill:#d1fae5"] .label {
color: #065f46 !important;
font-weight: 600 !important;
}
.mermaid .node[style*="fill:#f1f5f9"] rect,
.mermaid .node[style*="fill:#f1f5f9"] circle,
.mermaid .node[style*="fill:#f1f5f9"] polygon {
fill: #f1f5f9 !important;
stroke: #334155 !important;
stroke-width: 2px !important;
}
.mermaid .node[style*="fill:#f1f5f9"] .label {
color: #0f172a !important;
font-weight: 600 !important;
}
.mermaid .node[style*="fill:#fee2e2"] rect,
.mermaid .node[style*="fill:#fee2e2"] circle,
.mermaid .node[style*="fill:#fee2e2"] polygon {
fill: #fee2e2 !important;
stroke: #dc2626 !important;
stroke-width: 2px !important;
}
.mermaid .node[style*="fill:#fee2e2"] .label {
color: #991b1b !important;
font-weight: 600 !important;
}
.mermaid .node[style*="fill:#e0e7ff"] rect,
.mermaid .node[style*="fill:#e0e7ff"] circle,
.mermaid .node[style*="fill:#e0e7ff"] polygon {
fill: #e0e7ff !important;
stroke: #4338ca !important;
stroke-width: 2px !important;
}
.mermaid .node[style*="fill:#e0e7ff"] .label {
color: #3730a3 !important;
font-weight: 600 !important;
}
.mermaid .node[style*="fill:#f0f9ff"] rect,
.mermaid .node[style*="fill:#f0f9ff"] circle,
.mermaid .node[style*="fill:#f0f9ff"] polygon {
fill: #f0f9ff !important;
stroke: #0284c7 !important;
stroke-width: 2px !important;
}
.mermaid .node[style*="fill:#f0f9ff"] .label {
color: #0c4a6e !important;
font-weight: 600 !important;
}
.mermaid .node[style*="fill:#fdf2f8"] rect,
.mermaid .node[style*="fill:#fdf2f8"] circle,
.mermaid .node[style*="fill:#fdf2f8"] polygon {
fill: #fdf2f8 !important;
stroke: #be185d !important;
stroke-width: 2px !important;
}
.mermaid .node[style*="fill:#fdf2f8"] .label {
color: #831843 !important;
font-weight: 600 !important;
}
<span class="mention-invalid">@media</span> (max-width: 1024px) {
.mermaid-control-btn:not(.reset-zoom) {
display: none;
}
.mermaid-controls {
top: auto;
bottom: 15px;
right: 15px;
}
}
</style>
<base target="_blank">
</head>
<body class="bg-gradient-to-br from-stone-50 to-amber-50 text-stone-800 leading-relaxed">
<!-- Fixed Table of Contents -->
<nav id="toc-nav" class="fixed left-0 top-0 h-full w-80 bg-white/95 backdrop-blur-sm border-r border-stone-200 z-50 overflow-y-auto transform -translate-x-full lg:translate-x-0 transition-transform duration-300">
<div class="p-6">
<h3 class="text-lg font-semibold text-stone-900 mb-4 serif">目录</h3>
<ul class="space-y-2 text-sm">
<li>
<a href="#hero" class="block py-2 px-3 text-stone-600 hover:text-blue-600 hover:bg-stone-50 rounded transition-colors">引言</a>
</li>
<li>
<a href="#core-concepts" class="block py-2 px-3 text-stone-600 hover:text-blue-600 hover:bg-stone-50 rounded transition-colors">核心观点与主要论证</a>
</li>
<li>
<a href="#export-hypothesis" class="block py-2 px-3 text-stone-600 hover:text-blue-600 hover:bg-stone-50 rounded transition-colors">"出口假设"的深入剖析</a>
</li>
<li>
<a href="#neuroimaging" class="block py-2 px-3 text-stone-600 hover:text-blue-600 hover:bg-stone-50 rounded transition-colors">神经影像学证据支持</a>
</li>
<li>
<a href="#implications" class="block py-2 px-3 text-stone-600 hover:text-blue-600 hover:bg-stone-50 rounded transition-colors">对大模型机理的启示</a>
</li>
<li>
<a href="#brain-inspired" class="block py-2 px-3 text-stone-600 hover:text-blue-600 hover:bg-stone-50 rounded transition-colors">类脑智能研究启示</a>
</li>
<li>
<a href="#conclusion" class="block py-2 px-3 text-stone-600 hover:text-blue-600 hover:bg-stone-50 rounded transition-colors">总结与展望</a>
</li>
</ul>
</div>
</nav>
<!-- Mobile TOC Toggle -->
<button id="toc-toggle" class="lg:hidden fixed top-4 left-4 z-50 bg-white rounded-full p-3 shadow-lg">
<i class="fas fa-bars text-stone-600"></i>
</button>
<!-- Main Content -->
<div class="lg:ml-80">
<!-- Main Content Sections -->
<div class="max-w-6xl mx-auto px-4 sm:px-8 py-16 space-y-20">
<!-- Section 1: Core Concepts -->
<section id="core-concepts" class="space-y-12">
<header class="text-center space-y-4">
<h2 class="text-4xl font-light serif text-stone-900">核心观点与主要论证</h2>
<div class="w-24 h-1 bg-blue-600 mx-auto"></div>
<p class="text-xl text-stone-600 max-w-3xl mx-auto">
深入解析"出口假设"的理论框架，探索从浅层理解到深层理解的认知飞跃
</p>
</header>
<!-- Export Hypothesis Core -->
<div class="bg-white rounded-2xl p-8 shadow-lg border border-stone-200">
<h3 class="text-2xl font-semibold serif text-stone-900 mb-6">"出口假设"的核心论点</h3>
<div class="grid grid-cols-1 lg:grid-cols-2 gap-8">
<div class="space-y-4">
<p class="text-lg text-stone-700 leading-relaxed">
真正的语言理解需要将信息从核心语言系统<strong>"出口"（export）</strong>到其他功能特异化的脑区，以构建关于语言所描述情境的丰富心理模型。
<a href="https://medium.com/<span class="mention-invalid">@riazleghari</span>/why-ai-still-cant-truly-understand-language-b5931972e4ab" class="text-blue-600 hover:underline" target="_blank">
<sup>[48]</sup>
</a>
</p>
<div class="bg-stone-50 rounded-lg p-6">
<h4 class="font-semibold text-stone-900 mb-3">核心语言系统的局限性</h4>
<ul class="space-y-2 text-stone-700">
<li class="flex items-start">
<i class="fas fa-circle text-blue-500 text-xs mt-2 mr-3"></i>
<span>功能特异性：仅处理语言符号，不参与其他认知任务</span>
</li>
<li class="flex items-start">
<i class="fas fa-circle text-blue-500 text-xs mt-2 mr-3"></i>
<span>表征抽象性：脱离具体感知经验的符号操作</span>
</li>
<li class="flex items-start">
<i class="fas fa-circle text-blue-500 text-xs mt-2 mr-3"></i>
<span>统计规律性：基于语言共现关系的意义构建</span>
</li>
</ul>
</div>
</div>
<div class="space-y-4">
<img src="https://kimi-web-img.moonshot.cn/img/magicpic-p.cdn.bcebos.com/a30cea491125919ee440964cee8e97f572d5d030.png" alt="大脑语言区域与其他功能区域的神经连接示意图" class="w-full h-48 object-cover rounded-lg" size="medium" aspect="wide" query="大脑语言区域神经连接" referrerpolicy="no-referrer" data-modified="1" data-score="0.00"/>
<div class="bg-blue-50 rounded-lg p-4">
<h4 class="font-semibold text-blue-900 mb-2">邮件室隐喻</h4>
<p class="text-blue-800 text-sm">
核心语言系统如同一个邮件室，负责接收、分拣和标记信息，但真正的意义只有在这些信息被递送到正确的"部门"后才能产生。
<a href="https://medium.com/<span class="mention-invalid">@riazleghari</span>/why-ai-still-cant-truly-understand-language-b5931972e4ab" class="hover:underline" target="_blank">
<sup>[48]</sup>
</a>
</p>
</div>
</div>
</div>
</div>
<!-- Shallow vs Deep Understanding -->
<div class="grid grid-cols-1 md:grid-cols-2 gap-8">
<div class="bg-white rounded-2xl p-8 shadow-lg border border-stone-200">
<h3 class="text-xl font-semibold serif text-stone-900 mb-6 flex items-center">
<i class="fas fa-layer-group text-stone-500 mr-3"></i>
浅层理解
</h3>
<div class="space-y-4">
<p class="text-stone-700">
由大脑左半球额叶和颞叶区域构成的核心语言网络完成，功能局限于处理语言的表层形式。
<a href="https://medium.com/<span class="mention-invalid">@riazleghari</span>/why-ai-still-cant-truly-understand-language-b5931972e4ab" class="text-blue-600 hover:underline" target="_blank">
<sup>[48]</sup>
</a>
</p>
<div class="bg-stone-50 rounded-lg p-4">
<h4 class="font-semibold text-stone-900 mb-2">神经基础</h4>
<ul class="text-sm text-stone-600 space-y-1">
<li>• 左半球额叶语言区</li>
<li>• 颞叶语言网络</li>
<li>• 词汇识别与句法分析</li>
</ul>
</div>
<div class="bg-red-50 border-l-4 border-red-400 p-4">
<p class="text-red-800 text-sm">
<strong>局限性：</strong>停留在符号操作层面，与真实世界经验脱节
</p>
</div>
</div>
</div>
<div class="bg-white rounded-2xl p-8 shadow-lg border border-stone-200">
<h3 class="text-xl font-semibold serif text-stone-900 mb-6 flex items-center">
<i class="fas fa-brain text-blue-600 mr-3"></i>
深层理解
</h3>
<div class="space-y-4">
<p class="text-stone-700">
超越语言系统本身的全面认知体验，构建关于所描述情境的生动心理模型。
<a href="https://medium.com/<span class="mention-invalid">@riazleghari</span>/why-ai-still-cant-truly-understand-language-b5931972e4ab" class="text-blue-600 hover:underline" target="_blank">
<sup>[48]</sup>
</a>
</p>
<div class="bg-blue-50 rounded-lg p-4">
<h4 class="font-semibold text-blue-900 mb-2">整合过程</h4>
<ul class="text-sm text-blue-800 space-y-1">
<li>• 世界知识与自传体记忆</li>
<li>• 感知经验与运动表征</li>
<li>• 情感体验与情境模拟</li>
</ul>
</div>
<div class="bg-green-50 border-l-4 border-green-400 p-4">
<p class="text-green-800 text-sm">
<strong>特征：</strong>具身性（embodiment）和情境性（contextuality）的统一
</p>
</div>
</div>
</div>
</div>
<!-- Situation Models -->
<div class="bg-gradient-to-r from-stone-50 to-blue-50 rounded-2xl p-8 border border-stone-200">
<h3 class="text-2xl font-semibold serif text-stone-900 mb-6">情境模型（Situation Models）的概念</h3>
<div class="grid grid-cols-1 lg:grid-cols-3 gap-6">
<div class="lg:col-span-2 space-y-4">
<p class="text-stone-700 leading-relaxed">
情境模型是指个体在理解叙事性文本时，在脑海中构建的关于故事中人物、地点、事件及其相互关系的综合性表征。
<a href="https://smallake.kr/wp-content/uploads/2025/12/2511.19757v1.pdf" class="text-blue-600 hover:underline" target="_blank">
<sup>[41]</sup>
</a>
这种模型是动态的、多维度的，不仅包含文本明确提供的信息，还整合了个体的背景知识、推理和想象。
</p>
<blockquote class="bg-white rounded-lg p-6 border-l-4 border-blue-500">
<p class="text-stone-700 italic mb-2">
"当读到'Sid去了厨房'时，读者不仅会记住这个动作，还会推断Sid的意图（可能是为了找食物或喝水），并可能在脑海中构建一个关于厨房布局的模糊图像。"
</p>
<cite class="text-sm text-stone-500">— 情境模型构建示例</cite>
</blockquote>
</div>
<div class="space-y-4">
<img src="https://kimi-web-img.moonshot.cn/img/www.frontiersin.org/7a8ece7d5e580dd7c4f493be7223c9bd8f119ca5.jpg" alt="展示大脑处理语言时构建情境模型的示意图" class="w-full h-32 object-cover rounded-lg" size="small" aspect="wide" query="大脑语言理解情境模型" referrerpolicy="no-referrer" data-modified="1" data-score="0.00"/>
<div class="bg-white rounded-lg p-4">
<h4 class="font-semibold text-stone-900 mb-2">构建要素</h4>
<div class="space-y-2 text-sm">
<div class="flex items-center">
<i class="fas fa-user text-blue-500 mr-2"></i>
<span>人物角色与关系</span>
</div>
<div class="flex items-center">
<i class="fas fa-map-marker-alt text-green-500 mr-2"></i>
<span>空间场景与布局</span>
</div>
<div class="flex items-center">
<i class="fas fa-clock text-purple-500 mr-2"></i>
<span>时间序列与因果</span>
</div>
<div class="flex items-center">
<i class="fas fa-heart text-red-500 mr-2"></i>
<span>情感状态与动机</span>
</div>
</div>
</div>
</div>
</div>
</div>
</section>
<!-- Section 2: Export Hypothesis Analysis -->
<section id="export-hypothesis" class="space-y-12">
<header class="text-center space-y-4">
<h2 class="text-4xl font-light serif text-stone-900">"出口假设"的深入剖析</h2>
<div class="w-24 h-1 bg-purple-600 mx-auto"></div>
<p class="text-xl text-stone-600 max-w-3xl mx-auto">
探索信息导出的源头、目的地与本质过程，及其对AI研究范式的深远启示
</p>
</header>
<!-- Information Flow Diagram -->
<div class="bg-white rounded-2xl p-8 shadow-lg border border-stone-200">
<h3 class="text-2xl font-semibold serif text-stone-900 mb-6 text-center">信息导出流程</h3>
<div class="mermaid-container">
<div class="mermaid-controls">
<button class="mermaid-control-btn zoom-in" title="放大">
<i class="fas fa-search-plus"></i>
</button>
<button class="mermaid-control-btn zoom-out" title="缩小">
<i class="fas fa-search-minus"></i>
</button>
<button class="mermaid-control-btn reset-zoom" title="重置">
<i class="fas fa-expand-arrows-alt"></i>
</button>
<button class="mermaid-control-btn fullscreen" title="全屏查看">
<i class="fas fa-expand"></i>
</button>
</div>
<div class="mermaid" id="mermaid-diagram">
graph TD
A["语言输入
<br/>听觉/视觉"] --> B["核心语言系统
<br/>左半球额叶/颞叶"]
B --> C{"内容分析"}
C -->|"心理状态"| D["心理理论网络
<br/>rTPJ"]
C -->|"空间场景"| E["空间导航网络
<br/>PPA/RSC"]
C -->|"物理运动"| F["直觉物理网络
<br/>顶叶/前额叶"]
C -->|"感官描述"| G["感知皮层
<br/>视觉/听觉"]
C -->|"动作描述"| H["运动皮层
<br/>中央前回"]
D --> I["丰富心理模型"]
E --> I
F --> I
G --> I
H --> I
style A fill:#dbeafe,stroke:#1d4ed8,stroke-width:2px,color:#1e3a8a
style B fill:#f3f4f6,stroke:#4b5563,stroke-width:2px,color:#1f2937
style C fill:#fef3c7,stroke:#d97706,stroke-width:2px,color:#92400e
style D fill:#d1fae5,stroke:#059669,stroke-width:2px,color:#065f46
style E fill:#d1fae5,stroke:#059669,stroke-width:2px,color:#065f46
style F fill:#d1fae5,stroke:#059669,stroke-width:2px,color:#065f46
style G fill:#d1fae5,stroke:#059669,stroke-width:2px,color:#065f46
style H fill:#d1fae5,stroke:#059669,stroke-width:2px,color:#065f46
style I fill:#fef3c7,stroke:#d97706,stroke-width:3px,color:#92400e
</div>
</div>
<div class="mt-6 text-center">
<p class="text-sm text-stone-600">
<i class="fas fa-info-circle mr-1"></i>
支持缩放和拖拽操作，点击重置按钮恢复默认视图
</p>
</div>
</div>
<!-- Source, Destination, Process -->
<div class="grid grid-cols-1 lg:grid-cols-3 gap-8">
<div class="bg-white rounded-2xl p-8 shadow-lg border border-stone-200">
<div class="text-center mb-6">
<i class="fas fa-download text-blue-600 text-3xl mb-4"></i>
<h3 class="text-xl font-semibold serif text-stone-900">信息源点</h3>
</div>
<div class="space-y-4">
<h4 class="font-semibold text-stone-900">核心语言系统</h4>
<p class="text-stone-700 text-sm">
大脑左半球额叶和颞叶区域构成的专门化网络，负责处理语言信息的初步解码和结构化。
<a href="https://medium.com/<span class="mention-invalid">@riazleghari</span>/why-ai-still-cant-truly-understand-language-b5931972e4ab" class="text-blue-600 hover:underline" target="_blank">
<sup>[48]</sup>
</a>
</p>
<div class="bg-blue-50 rounded-lg p-4">
<h5 class="font-semibold text-blue-900 mb-2">功能特征</h5>
<ul class="text-sm text-blue-800 space-y-1">
<li>• 词汇识别与句法分析</li>
<li>• 抽象意义表征构建</li>
<li>• 语言统计规律提取</li>
<li>• 信息标准化处理</li>
</ul>
</div>
</div>
</div>
<div class="bg-white rounded-2xl p-8 shadow-lg border border-stone-200">
<div class="text-center mb-6">
<i class="fas fa-route text-green-600 text-3xl mb-4"></i>
<h3 class="text-xl font-semibold serif text-stone-900">导出过程</h3>
</div>
<div class="space-y-4">
<h4 class="font-semibold text-stone-900">符号到多模态转换</h4>
<p class="text-stone-700 text-sm">
从抽象符号处理到具体多模态表征的深刻转换，实现语言符号的"接地"过程。
<a href="https://medium.com/<span class="mention-invalid">@riazleghari</span>/why-ai-still-cant-truly-understand-language-b5931972e4ab" class="text-blue-600 hover:underline" target="_blank">
<sup>[48]</sup>
</a>
</p>
<div class="bg-green-50 rounded-lg p-4">
<h5 class="font-semibold text-green-900 mb-2">过程特征</h5>
<ul class="text-sm text-green-800 space-y-1">
<li>• 自动性与选择性</li>
<li>• 智能路由机制</li>
<li>• 跨模态信息整合</li>
<li>• 意义再情境化</li>
</ul>
</div>
</div>
</div>
<div class="bg-white rounded-2xl p-8 shadow-lg border border-stone-200">
<div class="text-center mb-6">
<i class="fas fa-bullseye text-purple-600 text-3xl mb-4"></i>
<h3 class="text-xl font-semibold serif text-stone-900">目的地</h3>
</div>
<div class="space-y-4">
<h4 class="font-semibold text-stone-900">专业脑区网络</h4>
<p class="text-stone-700 text-sm">
多个功能高度特化的脑区组成的分布式网络，各自负责处理和整合特定类型的知识和经验。
<a href="https://arxiv.org/abs/2511.19757" class="text-blue-600 hover:underline" target="_blank">
<sup>[36]</sup>
</a>
</p>
<div class="bg-purple-50 rounded-lg p-4">
<h5 class="font-semibold text-purple-900 mb-2">主要目的地</h5>
<ul class="text-sm text-purple-800 space-y-1">
<li>• 心理理论网络（rTPJ）</li>
<li>• 空间导航网络（PPA/RSC）</li>
<li>• 直觉物理网络</li>
<li>• 感知与运动皮层</li>
</ul>
</div>
</div>
</div>
</div>
<!-- AI Research Implications -->
<div class="bg-gradient-to-br from-blue-50 to-purple-50 rounded-2xl p-8 border border-stone-200">
<h3 class="text-2xl font-semibold serif text-stone-900 mb-6">对AI研究范式的启示</h3>
<div class="grid grid-cols-1 md:grid-cols-2 gap-8">
<div class="space-y-6">
<div class="bg-white rounded-lg p-6 shadow-sm">
<h4 class="font-semibold text-stone-900 mb-3 flex items-center">
<i class="fas fa-eye text-blue-600 mr-2"></i>
超越纯文本模型
</h4>
<p class="text-stone-700 text-sm mb-3">
真正的理解需要融合多模态信息和世界知识，构建具有感知经验的AI系统。
</p>
<div class="text-xs text-blue-600">
<a href="https://medium.com/<span class="mention-invalid">@riazleghari</span>/why-ai-still-cant-truly-understand-language-b5931972e4ab" class="hover:underline" target="_blank">
[48] 出口假设对AI的启示
</a>
</div>
</div>
<div class="bg-white rounded-lg p-6 shadow-sm">
<h4 class="font-semibold text-stone-900 mb-3 flex items-center">
<i class="fas fa-cubes text-green-600 mr-2"></i>
模块化架构设计
</h4>
<p class="text-stone-700 text-sm mb-3">
借鉴大脑功能分区思想，构建由专用子系统组成的松耦合AI架构。
</p>
<div class="text-xs text-green-600">
<a href="https://arxiv.org/abs/2511.19757" class="hover:underline" target="_blank">
[36] 神经科学基础研究
</a>
</div>
</div>
</div>
<div class="space-y-6">
<div class="bg-white rounded-lg p-6 shadow-sm">
<h4 class="font-semibold text-stone-900 mb-3 flex items-center">
<i class="fas fa-globe text-purple-600 mr-2"></i>
从预测到建模
</h4>
<p class="text-stone-700 text-sm mb-3">
从"预测下一个词"转向"构建世界模型"的目标转变。
</p>
<div class="text-xs text-purple-600">
<a href="https://www.pnas.org/doi/10.1073/pnas.2105646118" class="hover:underline" target="_blank">
[58] 语言模型训练目标分析
</a>
</div>
</div>
<div class="bg-gradient-to-r from-amber-100 to-orange-100 rounded-lg p-6">
<h4 class="font-semibold text-amber-900 mb-3">智人HRM模型启示</h4>
<p class="text-amber-800 text-sm">
清华大学万格智能团队参照人脑"分层—循环"机制搭建的类脑架构，体现了模块化设计的先进理念。
<a href="https://www.tsinghua.edu.cn/info/1182/121813.htm" class="text-amber-700 hover:underline" target="_blank">
<sup>[63]</sup>
</a>
</p>
</div>
</div>
</div>
</div>
</section>
<!-- Section 3: Neuroimaging Evidence -->
<section id="neuroimaging" class="space-y-12">
<header class="text-center space-y-4">
<h2 class="text-4xl font-light serif text-stone-900">神经影像学证据支持</h2>
<div class="w-24 h-1 bg-green-600 mx-auto"></div>
<p class="text-xl text-stone-600 max-w-3xl mx-auto">
fMRI技术为"出口假设"提供坚实的实证基础，揭示语言理解的神经机制
</p>
</header>
<!-- fMRI Key Role -->
<div class="bg-white rounded-2xl p-8 shadow-lg border border-stone-200">
<h3 class="text-2xl font-semibold serif text-stone-900 mb-6">功能性磁共振成像的关键作用</h3>
<div class="grid grid-cols-1 lg:grid-cols-3 gap-6">
<div class="space-y-4">
<div class="bg-blue-50 rounded-lg p-6 text-center">
<i class="fas fa-map-marker-alt text-blue-600 text-3xl mb-4"></i>
<h4 class="font-semibold text-blue-900 mb-2">定位功能</h4>
<p class="text-blue-800 text-sm">
精确描绘核心语言系统轮廓，识别语言特异性激活脑区
</p>
</div>
</div>
<div class="space-y-4">
<div class="bg-green-50 rounded-lg p-6 text-center">
<i class="fas fa-network-wired text-green-600 text-3xl mb-4"></i>
<h4 class="font-semibold text-green-900 mb-2">连接功能</h4>
<p class="text-green-800 text-sm">
识别信息导出的目的地脑区，揭示跨系统协同模式
</p>
</div>
</div>
<div class="space-y-4">
<div class="bg-purple-50 rounded-lg p-6 text-center">
<i class="fas fa-chart-line text-purple-600 text-3xl mb-4"></i>
<h4 class="font-semibold text-purple-900 mb-2">动态监测</h4>
<p class="text-purple-800 text-sm">
实时观测不同脑区的协同激活模式和信息交换过程
</p>
</div>
</div>
</div>
</div>
<!-- Evidence Chain -->
<div class="space-y-8">
<h3 class="text-2xl font-semibold serif text-stone-900 text-center">证据链：特定内容激活特定脑区</h3>
<!-- Spatial Navigation Evidence -->
<div class="bg-white rounded-2xl p-8 shadow-lg border border-stone-200">
<h4 class="text-xl font-semibold text-stone-900 mb-6 flex items-center">
<i class="fas fa-map text-blue-600 mr-3"></i>
空间与导航描述激活场景理解相关脑区
</h4>
<div class="grid grid-cols-1 lg:grid-cols-2 gap-8">
<div class="space-y-4">
<p class="text-stone-700">
当语言描述涉及地点、空间布局或导航路径时，大脑中负责场景理解和空间导航的特定区域会被显著激活，主要包括<strong>旁海马回位置区（PPA）和压后皮层（RSC）</strong>。
<a href="https://medium.com/<span class="mention-invalid">@riazleghari</span>/why-ai-still-cant-truly-understand-language-b5931972e4ab" class="text-blue-600 hover:underline" target="_blank">
<sup>[48]</sup>
</a>
</p>
<div class="bg-blue-50 rounded-lg p-4">
<h5 class="font-semibold text-blue-900 mb-2">关键发现</h5>
<ul class="text-sm text-blue-800 space-y-1">
<li>• 文字描述即可激活PPA和RSC</li>
<li>• 激活具有高度选择性</li>
<li>• 构建类似视觉体验的心理模型</li>
<li>• 场景描述比物体描述激活更强</li>
</ul>
</div>
</div>
<div class="space-y-4">
<img src="https://fixedplaceholder" alt="大脑海马旁回位置区域（PPA）和压后皮层（RSC）的解剖结构示意图" class="w-full h-40 object-cover rounded-lg" size="medium" aspect="wide" query="大脑海马旁回位置区和压后皮层解剖图" referrerpolicy="no-referrer" data-modified="1" data-score="0.00"/>
<div class="bg-stone-100 rounded-lg p-4">
<h5 class="font-semibold text-stone-900 mb-2">脑区功能</h5>
<div class="space-y-2 text-sm">
<div class="flex justify-between">
<span class="text-stone-600">PPA</span>
<span class="text-stone-700">场景识别与处理</span>
</div>
<div class="flex justify-between">
<span class="text-stone-600">RSC</span>
<span class="text-stone-700">空间导航与记忆</span>
</div>
</div>
</div>
</div>
</div>
</div>
<!-- Theory of Mind Evidence -->
<div class="bg-white rounded-2xl p-8 shadow-lg border border-stone-200">
<h4 class="text-xl font-semibold text-stone-900 mb-6 flex items-center">
<i class="fas fa-users text-green-600 mr-3"></i>
涉及他人心理状态的描述激活"心理理论"网络
</h4>
<div class="grid grid-cols-1 lg:grid-cols-2 gap-8">
<div class="space-y-4">
<p class="text-stone-700">
当语言涉及他人的信念、意图、欲望等心理状态时，信息会被传递到以<strong>右侧颞顶交界区（rTPJ）</strong>为核心的ToM网络，这个网络专门负责推断和理解他人的内心世界。
<a href="https://arxiv.org/html/2511.19757v1" class="text-blue-600 hover:underline" target="_blank">
<sup>[45]</sup>
</a>
</p>
<blockquote class="bg-green-50 border-l-4 border-green-400 p-4">
<p class="text-green-800 text-sm italic">
"她以为他爱她，但他心里却想着另一个人"
</p>
<cite class="text-green-600 text-xs">— 涉及复杂心理状态的句子示例</cite>
</blockquote>
<div class="bg-green-50 rounded-lg p-4">
<h5 class="font-semibold text-green-900 mb-2">关键证据</h5>
<ul class="text-sm text-green-800 space-y-1">
<li>• 被动听故事时自动激活</li>
<li>• 激活程度与语言系统独立</li>
<li>• 专门处理心理状态信息</li>
<li>• 社会认知的关键脑区</li>
</ul>
</div>
</div>
<div class="space-y-4">
<img src="https://fixedplaceholder" alt="大脑右侧颞顶交界区（rTPJ）位置示意图" class="w-full h-40 object-cover rounded-lg" size="medium" aspect="wide" style="linedrawing" query="大脑右侧颞顶交界区解剖位置" referrerpolicy="no-referrer" data-modified="1" data-score="0.00"/>
<div class="bg-stone-100 rounded-lg p-4">
<h5 class="font-semibold text-stone-900 mb-2">ToM网络组成</h5>
<div class="space-y-2 text-sm">
<div class="flex items-center justify-between">
<span class="text-stone-600">rTPJ</span>
<span class="text-stone-700 bg-red-100 px-2 py-1 rounded">核心节点</span>
</div>
<div class="flex items-center justify-between">
<span class="text-stone-600">前额叶皮层</span>
<span class="text-stone-700 bg-blue-100 px-2 py-1 rounded">支持区域</span>
</div>
<div class="flex items-center justify-between">
<span class="text-stone-600">颞上沟</span>
<span class="text-stone-700 bg-green-100 px-2 py-1 rounded">辅助区域</span>
</div>
</div>
</div>
</div>
</div>
</div>
<!-- Sensory-Motor Evidence -->
<div class="bg-white rounded-2xl p-8 shadow-lg border border-stone-200">
<h4 class="text-xl font-semibold text-stone-900 mb-6 flex items-center">
<i class="fas fa-hand-paper text-purple-600 mr-3"></i>
生动具体的描述激活感知与运动皮层
</h4>
<div class="grid grid-cols-1 lg:grid-cols-3 gap-6">
<div class="space-y-4">
<p class="text-stone-700 text-sm">
当语言描述涉及具体的动作或生动的感官体验时，相应的运动皮层和感知皮层会被激活，这种现象被称为<strong>"动作-语义一致性效应"</strong>。
<a href="https://medium.com/<span class="mention-invalid">@riazleghari</span>/why-ai-still-cant-truly-understand-language-b5931972e4ab" class="text-blue-600 hover:underline" target="_blank">
<sup>[48]</sup>
</a>
</p>
<div class="bg-purple-50 rounded-lg p-4">
<h5 class="font-semibold text-purple-900 mb-2">具身性理解</h5>
<p class="text-purple-800 text-xs">
通过"感知模拟"重现相关经验，使理解变得生动具体
</p>
</div>
</div>
<div class="space-y-4">
<h5 class="font-semibold text-stone-900">动作描述激活</h5>
<div class="space-y-2 text-sm">
<div class="flex items-center">
<i class="fas fa-hand-rock text-blue-500 mr-2"></i>
<span class="text-stone-700">"抓"、"扔" → 手部运动区</span>
</div>
<div class="flex items-center">
<i class="fas fa-running text-green-500 mr-2"></i>
<span class="text-stone-700">"跑"、"跳" → 腿部运动区</span>
</div>
<div class="flex items-center">
<i class="fas fa-smile text-yellow-500 mr-2"></i>
<span class="text-stone-700">面部表情 → 面部运动区</span>
</div>
</div>
</div>
<div class="space-y-4">
<h5 class="font-semibold text-stone-900">感官描述激活</h5>
<div class="space-y-2 text-sm">
<div class="flex items-center">
<i class="fas fa-eye text-red-500 mr-2"></i>
<span class="text-stone-700">"闪闪发光" → 视觉皮层</span>
</div>
<div class="flex items-center">
<i class="fas fa-hand-point-up text-orange-500 mr-2"></i>
<span class="text-stone-700">"粗糙" → 体感皮层</span>
</div>
<div class="flex items-center">
<i class="fas fa-volume-up text-purple-500 mr-2"></i>
<span class="text-stone-700">"响亮" → 听觉皮层</span>
</div>
</div>
</div>
</div>
</div>
</div>
<!-- Theoretical Framework Visualization -->
<div class="bg-gradient-to-br from-stone-50 to-blue-50 rounded-2xl p-8 border border-stone-200">
<h3 class="text-2xl font-semibold serif text-stone-900 mb-6 text-center">理论框架的可视化呈现</h3>
<div class="grid grid-cols-1 lg:grid-cols-2 gap-8">
<div class="space-y-6">
<div class="bg-white rounded-lg p-6 shadow-sm">
<h4 class="font-semibold text-stone-900 mb-4 flex items-center">
<i class="fas fa-project-diagram text-blue-600 mr-2"></i>
信息导出路径
</h4>
<p class="text-stone-700 text-sm mb-4">
图示清晰展示语言理解作为分布式、多系统协同的过程，强调核心语言系统作为信息枢纽的角色。
</p>
<div class="space-y-2 text-xs">
<div class="flex items-center">
<div class="w-3 h-3 bg-blue-500 rounded-full mr-2"></div>
<span class="text-stone-600">核心语言系统（蓝色）</span>
</div>
<div class="flex items-center">
<div class="w-3 h-3 bg-green-500 rounded-full mr-2"></div>
<span class="text-stone-600">专业处理系统（多色）</span>
</div>
<div class="flex items-center">
<div class="w-3 h-3 bg-yellow-500 rounded-full mr-2"></div>
<span class="text-stone-600">丰富心理模型（黄色）</span>
</div>
</div>
</div>
</div>
<div class="space-y-6">
<div class="bg-white rounded-lg p-6 shadow-sm">
<h4 class="font-semibold text-stone-900 mb-4 flex items-center">
<i class="fas fa-balance-scale text-purple-600 mr-2"></i>
理解层次对比
</h4>
<div class="space-y-4">
<div class="border-l-4 border-stone-400 pl-4">
<h5 class="font-semibold text-stone-900 text-sm">浅层理解</h5>
<p class="text-stone-600 text-xs">仅核心语言系统激活，局限于符号处理</p>
</div>
<div class="border-l-4 border-blue-500 pl-4">
<h5 class="font-semibold text-blue-900 text-sm">深层理解</h5>
<p class="text-blue-800 text-xs">多系统协同激活，构建丰富心理模型</p>
</div>
</div>
</div>
<div class="bg-stone-100 rounded-lg p-4">
<h5 class="font-semibold text-stone-900 mb-2">神经资源调动</h5>
<p class="text-stone-600 text-xs">
深层理解需要调动全脑范围的神经资源，实现从符号操作到意义生成的飞跃
</p>
</div>
</div>
</div>
</div>
</section>
<!-- Section 4: LLM Implications -->
<section id="implications" class="space-y-12">
<header class="text-center space-y-4">
<h2 class="text-4xl font-light serif text-stone-900">对大模型机理的启示</h2>
<div class="w-24 h-1 bg-red-600 mx-auto"></div>
<p class="text-xl text-stone-600 max-w-3xl mx-auto">
从神经科学视角审视大语言模型的理解瓶颈与改进路径
</p>
</header>
<!-- LLM Understanding Bottleneck -->
<div class="bg-white rounded-2xl p-8 shadow-lg border border-stone-200">
<h3 class="text-2xl font-semibold serif text-stone-900 mb-6">解释LLM的"理解"瓶颈</h3>
<div class="grid grid-cols-1 lg:grid-cols-2 gap-8">
<div class="space-y-6">
<div class="bg-red-50 border-l-4 border-red-400 p-6">
<h4 class="font-semibold text-red-900 mb-3 flex items-center">
<i class="fas fa-exclamation-triangle text-red-600 mr-2"></i>
根本瓶颈
</h4>
<p class="text-red-800 text-sm">
LLM缺乏与真实世界相连的"出口"，整个认知世界构建在文本数据之上，无法将语言符号与真实世界的感知、运动和情感体验联系起来。
<a href="https://medium.com/<span class="mention-invalid">@riazleghari</span>/why-ai-still-cant-truly-understand-language-b5931972e4ab" class="hover:underline" target="_blank">
<sup>[48]</sup>
</a>
</p>
</div>
<div class="bg-stone-50 rounded-lg p-6">
<h4 class="font-semibold text-stone-900 mb-3">类比分析</h4>
<p class="text-stone-700 text-sm">
LLM就像一个被关在图书馆里、从未见过外面世界的博学之士。它知道所有关于"猫"的词汇和描述，但从未真正"见过"猫、"摸过"猫，也从未体验过与猫互动的情感。
</p>
</div>
</div>
<div class="space-y-6">
<img src="https://fixedplaceholder" alt="被书籍包围的机器人形象" class="w-full h-48 object-cover rounded-lg" size="medium" aspect="wide" style="photo" query="机器人 图书馆 书籍" referrerpolicy="no-referrer" data-modified="1" data-score="0.00"/>
<div class="bg-amber-50 rounded-lg p-6">
<h4 class="font-semibold text-amber-900 mb-3">局限性表现</h4>
<div class="space-y-2 text-sm text-amber-800">
<div class="flex items-center">
<i class="fas fa-circle text-xs mr-2"></i>
<span>产生"幻觉"（hallucination）</span>
</div>
<div class="flex items-center">
<i class="fas fa-circle text-xs mr-2"></i>
<span>不合逻辑的推理</span>
</div>
<div class="flex items-center">
<i class="fas fa-circle text-xs mr-2"></i>
<span>缺乏常识性理解</span>
</div>
<div class="flex items-center">
<i class="fas fa-circle text-xs mr-2"></i>
<span>符号层面的"去情境化"</span>
</div>
</div>
</div>
</div>
</div>
</div>
<!-- World Model Construction -->
<div class="grid grid-cols-1 lg:grid-cols-2 gap-8">
<div class="bg-white rounded-2xl p-8 shadow-lg border border-stone-200">
<h3 class="text-xl font-semibold serif text-stone-900 mb-6">构建"世界模型"的可行路径</h3>
<div class="space-y-6">
<div class="bg-blue-50 rounded-lg p-6">
<h4 class="font-semibold text-blue-900 mb-3 flex items-center">
<i class="fas fa-eye text-blue-600 mr-2"></i>
多模态融合
</h4>
<p class="text-blue-800 text-sm mb-3">
让模型同时处理文本、图像、声音、视频甚至机器人传感器数据，为其提供丰富的多模态信息。
</p>
<div class="flex flex-wrap gap-2">
<span class="bg-blue-200 text-blue-800 px-2 py-1 rounded text-xs">视觉</span>
<span class="bg-blue-200 text-blue-800 px-2 py-1 rounded text-xs">听觉</span>
<span class="bg-blue-200 text-blue-800 px-2 py-1 rounded text-xs">触觉</span>
<span class="bg-blue-200 text-blue-800 px-2 py-1 rounded text-xs">传感器</span>
</div>
</div>
<div class="bg-green-50 rounded-lg p-6">
<h4 class="font-semibold text-green-900 mb-3 flex items-center">
<i class="fas fa-database text-green-600 mr-2"></i>
结构化知识
</h4>
<p class="text-green-800 text-sm mb-3">
整合知识图谱、物理引擎、符号逻辑系统等结构化知识库。
</p>
<div class="flex flex-wrap gap-2">
<span class="bg-green-200 text-green-800 px-2 py-1 rounded text-xs">知识图谱</span>
<span class="bg-green-200 text-green-800 px-2 py-1 rounded text-xs">物理引擎</span>
<span class="bg-green-200 text-green-800 px-2 py-1 rounded text-xs">符号逻辑</span>
</div>
</div>
<div class="bg-purple-50 rounded-lg p-6">
<h4 class="font-semibold text-purple-900 mb-3 flex items-center">
<i class="fas fa-robot text-purple-600 mr-2"></i>
具身智能
</h4>
<p class="text-purple-800 text-sm">
将AI模型置于物理实体中，通过传感器感知环境，在与真实世界的直接互动中学习和构建世界模型。
</p>
</div>
</div>
</div>
<div class="bg-white rounded-2xl p-8 shadow-lg border border-stone-200">
<h3 class="text-xl font-semibold serif text-stone-900 mb-6">表征相似性分析框架</h3>
<div class="space-y-6">
<img src="https://kimi-web-img.moonshot.cn/img/deepaiedu.com/4fd2574c609948db1e6cc03c86bf7270b165eda8.png" alt="大脑神经活动与人工智能模型表征的对比示意图" class="w-full h-32 object-cover rounded-lg" size="small" aspect="wide" query="神经科学与人工智能对比" referrerpolicy="no-referrer" data-modified="1" data-score="0.00"/>
<div class="bg-stone-50 rounded-lg p-6">
<h4 class="font-semibold text-stone-900 mb-4">比较维度</h4>
<div class="space-y-3">
<div class="flex items-start">
<i class="fas fa-chart-bar text-blue-500 mt-1 mr-3"></i>
<div>
<h5 class="font-semibold text-stone-900 text-sm">表征相似性分析（RSA）</h5>
<p class="text-stone-600 text-xs">比较LLM内部激活与人类fMRI数据的表征结构</p>
</div>
</div>
<div class="flex items-start">
<i class="fas fa-search text-green-500 mt-1 mr-3"></i>
<div>
<h5 class="font-semibold text-stone-900 text-sm">"出口"机制探测</h5>
<p class="text-stone-600 text-xs">检测模型是否存在类似人类的模块激活模式</p>
</div>
</div>
<div class="flex items-start">
<i class="fas fa-cogs text-purple-500 mt-1 mr-3"></i>
<div>
<h5 class="font-semibold text-stone-900 text-sm">因果干预实验</h5>
<p class="text-stone-600 text-xs">通过干预内部状态探究不同模块的功能</p>
</div>
</div>
</div>
</div>
<div class="bg-amber-50 rounded-lg p-4">
<p class="text-amber-800 text-sm">
<i class="fas fa-lightbulb mr-2"></i>
通过这些精细的比较，可以客观判断LLM的"理解"究竟停留在"浅层"符号操作，还是触及了"深层"的、与世界知识相关联的表征。
</p>
</div>
</div>
</div>
</div>
</section>
<!-- Section 5: Brain-Inspired AI -->
<section id="brain-inspired" class="space-y-12">
<header class="text-center space-y-4">
<h2 class="text-4xl font-light serif text-stone-900">类脑智能研究启示</h2>
<div class="w-24 h-1 bg-indigo-600 mx-auto"></div>
<p class="text-xl text-stone-600 max-w-3xl mx-auto">
从神经科学发现中汲取灵感，构建更"类人"的智能系统
</p>
</header>
<!-- Cognitive Architecture Design -->
<div class="bg-white rounded-2xl p-8 shadow-lg border border-stone-200">
<h3 class="text-2xl font-semibold serif text-stone-900 mb-6">设计更"类人"的认知架构</h3>
<div class="grid grid-cols-1 lg:grid-cols-2 gap-8">
<div class="space-y-6">
<div class="bg-blue-50 rounded-lg p-6">
<h4 class="font-semibold text-blue-900 mb-3 flex items-center">
<i class="fas fa-puzzle-piece text-blue-600 mr-2"></i>
模块化设计优势
</h4>
<div class="space-y-2 text-sm text-blue-800">
<div class="flex items-center">
<i class="fas fa-check-circle text-xs mr-2"></i>
<span>提高系统可解释性和可维护性</span>
</div>
<div class="flex items-center">
<i class="fas fa-check-circle text-xs mr-2"></i>
<span>独立模块优化和调试</span>
</div>
<div class="flex items-center">
<i class="fas fa-check-circle text-xs mr-2"></i>
<span>降低系统复杂性</span>
</div>
</div>
</div>
<div class="bg-green-50 rounded-lg p-6">
<h4 class="font-semibold text-green-900 mb-3 flex items-center">
<i class="fas fa-share-alt text-green-600 mr-2"></i>
分布式处理优势
</h4>
<div class="space-y-2 text-sm text-green-800">
<div class="flex items-center">
<i class="fas fa-check-circle text-xs mr-2"></i>
<span>并行处理提高效率</span>
</div>
<div class="flex items-center">
<i class="fas fa-check-circle text-xs mr-2"></i>
<span>增强系统鲁棒性</span>
</div>
<div class="flex items-center">
<i class="fas fa-check-circle text-xs mr-2"></i>
<span>功能代偿能力</span>
</div>
</div>
</div>
<div class="bg-purple-50 rounded-lg p-6">
<h4 class="font-semibold text-purple-900 mb-3 flex items-center">
<i class="fas fa-cogs text-purple-600 mr-2"></i>
功能特化优势
</h4>
<div class="space-y-2 text-sm text-purple-800">
<div class="flex items-center">
<i class="fas fa-check-circle text-xs mr-2"></i>
<span>深度优化特定任务</span>
</div>
<div class="flex items-center">
<i class="fas fa-check-circle text-xs mr-2"></i>
<span>专业化性能提升</span>
</div>
<div class="flex items-center">
<i class="fas fa-check-circle text-xs mr-2"></i>
<span>应对复杂多样挑战</span>
</div>
</div>
</div>
</div>
<div class="space-y-6">
<img src="https://kimi-web-img.moonshot.cn/img/psy.swu.edu.cn/3af74c5d78e63f6e8269baedeb356dc88af868eb.png" alt="人脑神经网络模块化结构示意图" class="w-full h-48 object-cover rounded-lg" size="medium" aspect="wide" query="人脑神经网络模块化结构" referrerpolicy="no-referrer" data-modified="1" data-score="0.00"/>
<div class="bg-gradient-to-r from-amber-100 to-orange-100 rounded-lg p-6">
<h4 class="font-semibold text-amber-900 mb-3">智人HRM模型案例</h4>
<p class="text-amber-800 text-sm mb-3">
清华大学万格智能团队研发的HRM模型，参照人脑"分层—循环"机制搭建类脑架构，体现了模块化设计的先进理念。
<a href="https://www.tsinghua.edu.cn/info/1182/121813.htm" class="text-amber-700 hover:underline" target="_blank">
<sup>[63]</sup>
</a>
</p>
<div class="flex flex-wrap gap-2">
<span class="bg-amber-200 text-amber-800 px-2 py-1 rounded text-xs">分层架构</span>
<span class="bg-amber-200 text-amber-800 px-2 py-1 rounded text-xs">循环机制</span>
<span class="bg-amber-200 text-amber-800 px-2 py-1 rounded text-xs">类脑设计</span>
</div>
</div>
</div>
</div>
</div>
<!-- Multi-modal Integration -->
<div class="grid grid-cols-1 lg:grid-cols-2 gap-8">
<div class="bg-white rounded-2xl p-8 shadow-lg border border-stone-200">
<h3 class="text-xl font-semibold serif text-stone-900 mb-6">多模态整合与常识推理</h3>
<div class="space-y-6">
<div class="bg-gradient-to-r from-blue-50 to-purple-50 rounded-lg p-6">
<h4 class="font-semibold text-stone-900 mb-4">具身心智（Embodied Mind）</h4>
<p class="text-stone-700 text-sm mb-4">
真正智能的体不应仅仅是语言处理机，而应像人类一样通过多种感官感知世界，并运用常识进行思考和行动。
</p>
<div class="grid grid-cols-2 gap-4">
<div class="text-center">
<i class="fas fa-eye text-blue-600 text-2xl mb-2"></i>
<p class="text-xs text-stone-600">视觉感知</p>
</div>
<div class="text-center">
<i class="fas fa-ear text-green-600 text-2xl mb-2"></i>
<p class="text-xs text-stone-600">听觉感知</p>
</div>
<div class="text-center">
<i class="fas fa-hand-paper text-purple-600 text-2xl mb-2"></i>
<p class="text-xs text-stone-600">触觉感知</p>
</div>
<div class="text-center">
<i class="fas fa-brain text-indigo-600 text-2xl mb-2"></i>
<p class="text-xs text-stone-600">认知推理</p>
</div>
</div>
</div>
<div class="space-y-4">
<div class="bg-blue-50 rounded-lg p-4">
<h5 class="font-semibold text-blue-900 mb-2">多模态表征学习</h5>
<p class="text-blue-800 text-sm">
有效融合不同模态信息，形成统一连贯的内部表征
</p>
</div>
<div class="bg-green-50 rounded-lg p-4">
<h5 class="font-semibold text-green-900 mb-2">世界模型构建</h5>
<p class="text-green-800 text-sm">
在与环境交互中主动学习和构建动态世界模型
</p>
</div>
<div class="bg-purple-50 rounded-lg p-4">
<h5 class="font-semibold text-purple-900 mb-2">常识推理实现</h5>
<p class="text-purple-800 text-sm">
结合知识库与神经网络，实现灵活的常识推理
</p>
</div>
</div>
</div>
</div>
<div class="bg-white rounded-2xl p-8 shadow-lg border border-stone-200">
<h3 class="text-xl font-semibold serif text-stone-900 mb-6">学习范式转变</h3>
<div class="space-y-6">
<div class="bg-stone-50 rounded-lg p-6">
<h4 class="font-semibold text-stone-900 mb-4">被动吸收 vs 主动构建</h4>
<div class="space-y-4">
<div class="border-l-4 border-stone-400 pl-4">
<h5 class="font-semibold text-stone-900 text-sm">当前范式：被动吸收</h5>
<ul class="text-stone-600 text-xs mt-1 space-y-1">
<li>• 监督式学习</li>
<li>• 静态数据集训练</li>
<li>• 无内在目标驱动</li>
</ul>
</div>
<div class="border-l-4 border-blue-500 pl-4">
<h5 class="font-semibold text-blue-900 text-sm">未来范式：主动构建</h5>
<ul class="text-blue-800 text-xs mt-1 space-y-1">
<li>• 内在动机与好奇心</li>
<li>• 基于模型的强化学习</li>
<li>• 持续学习与终身学习</li>
</ul>
</div>
</div>
</div>
<div class="bg-gradient-to-r from-indigo-50 to-blue-50 rounded-lg p-6">
<h4 class="font-semibold text-indigo-900 mb-4">主动学习特征</h4>
<div class="space-y-3">
<div class="flex items-start">
<i class="fas fa-question-circle text-indigo-600 text-sm mt-1 mr-3"></i>
<div>
<h5 class="font-semibold text-indigo-900 text-sm">内在好奇心</h5>
<p class="text-indigo-800 text-xs">主动探索未知，寻求新奇体验</p>
</div>
</div>
<div class="flex items-start">
<i class="fas fa-brain text-indigo-600 text-sm mt-1 mr-3"></i>
<div>
<h5 class="font-semibold text-indigo-900 text-sm">心理模拟</h5>
<p class="text-indigo-800 text-xs">基于世界模型进行预测和规划</p>
</div>
</div>
<div class="flex items-start">
<i class="fas fa-sync text-indigo-600 text-sm mt-1 mr-3"></i>
<div>
<h5 class="font-semibold text-indigo-900 text-sm">持续适应</h5>
<p class="text-indigo-800 text-xs">在不断变化的环境中学习和成长</p>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</section>
<!-- Conclusion -->
<section id="conclusion" class="space-y-12">
<header class="text-center space-y-4">
<h2 class="text-4xl font-light serif text-stone-900">总结与展望</h2>
<div class="w-24 h-1 bg-amber-600 mx-auto"></div>
<p class="text-xl text-stone-600 max-w-3xl mx-auto">
"出口假设"为AI研究指明方向：从语言模仿到真正理解的认知飞跃
</p>
</header>
<!-- Key Insights Summary -->
<div class="bg-gradient-to-br from-stone-100 to-amber-50 rounded-2xl p-8 border border-stone-200">
<div class="grid grid-cols-1 lg:grid-cols-2 gap-8">
<div class="space-y-6">
<h3 class="text-2xl font-semibold serif text-stone-900 mb-4">核心发现总结</h3>
<div class="space-y-4">
<div class="bg-white rounded-lg p-6 shadow-sm">
<h4 class="font-semibold text-stone-900 mb-2 flex items-center">
<i class="fas fa-brain text-blue-600 mr-2"></i>
神经科学洞察
</h4>
<p class="text-stone-700 text-sm">
真正的语言理解需要将信息从核心语言系统"出口"到其他功能特异化的脑区，构建丰富的心理模型。
</p>
</div>
<div class="bg-white rounded-lg p-6 shadow-sm">
<h4 class="font-semibold text-stone-900 mb-2 flex items-center">
<i class="fas fa-robot text-green-600 mr-2"></i>
AI瓶颈识别
</h4>
<p class="text-stone-700 text-sm">
大语言模型的根本局限在于缺乏与真实世界的"出口"连接，其理解停留在符号操作层面。
</p>
</div>
<div class="bg-white rounded-lg p-6 shadow-sm">
<h4 class="font-semibold text-stone-900 mb-2 flex items-center">
<i class="fas fa-lightbulb text-purple-600 mr-2"></i>
未来方向
</h4>
<p class="text-stone-700 text-sm">
构建多模态融合、模块化架构、世界模型导向的新一代AI系统。
</p>
</div>
</div>
</div>
<div class="space-y-6">
<img src="https://kimi-web-img.moonshot.cn/img/www.forwardpathway.com/ce58c999f981070cc9ddee78acd2c75f8cbfab7f.jpg" alt="人工智能与人类大脑融合的未来科技概念图" class="w-full h-48 object-cover rounded-lg" size="medium" aspect="wide" query="人工智能与大脑融合的未来概念图" referrerpolicy="no-referrer" data-modified="1" data-score="0.00"/>
<div class="bg-amber-50 rounded-lg p-6">
<h4 class="font-semibold text-amber-900 mb-4">研究意义</h4>
<div class="space-y-2 text-sm text-amber-800">
<div class="flex items-center">
<i class="fas fa-microscope text-xs mr-2"></i>
<span>为认知神经科学提供新视角</span>
</div>
<div class="flex items-center">
<i class="fas fa-cogs text-xs mr-2"></i>
<span>为AI架构设计指明方向</span>
</div>
<div class="flex items-center">
<i class="fas fa-bridge text-xs mr-2"></i>
<span>构建跨学科研究桥梁</span>
</div>
<div class="flex items-center">
<i class="fas fa-rocket text-xs mr-2"></i>
<span>推动通用人工智能发展</span>
</div>
</div>
</div>
</div>
</div>
</div>
<!-- Future Research Directions -->
<div class="bg-white rounded-2xl p-8 shadow-lg border border-stone-200">
<h3 class="text-2xl font-semibold serif text-stone-900 mb-6 text-center">未来研究方向</h3>
<div class="grid grid-cols-1 md:grid-cols-3 gap-6">
<div class="bg-blue-50 rounded-lg p-6 text-center">
<i class="fas fa-puzzle-piece text-blue-600 text-3xl mb-4"></i>
<h4 class="font-semibold text-blue-900 mb-3">认知架构优化</h4>
<p class="text-blue-800 text-sm">
基于大脑模块化思想，设计更高效的分布式AI架构，实现专业化与协同的统一
</p>
</div>
<div class="bg-green-50 rounded-lg p-6 text-center">
<i class="fas fa-globe text-green-600 text-3xl mb-4"></i>
<h4 class="font-semibold text-green-900 mb-3">世界模型构建</h4>
<p class="text-green-800 text-sm">
探索多模态融合与具身智能路径，构建真正理解世界的AI系统
</p>
</div>
<div class="bg-purple-50 rounded-lg p-6 text-center">
<i class="fas fa-sync-alt text-purple-600 text-3xl mb-4"></i>
<h4 class="font-semibold text-purple-900 mb-3">学习范式革新</h4>
<p class="text-purple-800 text-sm">
从被动数据吸收转向主动构建心理模型，实现真正的智能学习
</p>
</div>
</div>
</div>
<!-- Final Reflection -->
<div class="bg-gradient-to-r from-stone-900 to-blue-900 rounded-2xl p-8 text-white">
<div class="text-center space-y-6">
<i class="fas fa-quote-left text-4xl opacity-50"></i>
<blockquote class="text-2xl font-light serif italic leading-relaxed">
"从'预测下一个词'到'构建世界模型'的转变，代表了AI研究从追求语言模仿能力到追求真正认知能力的范式转移，是实现通用人工智能的必经之路。"
</blockquote>
<div class="w-16 h-1 bg-amber-400 mx-auto"></div>
<p class="text-amber-200 text-lg">
— 基于"出口假设"的未来展望
</p>
</div>
</div>
</section>
</div>
<!-- References -->
<div class="max-w-6xl mx-auto px-4 sm:px-8 py-16 border-t border-stone-200">
<h3 class="text-2xl font-semibold serif text-stone-900 mb-8">参考文献</h3>
<div class="grid grid-cols-1 md:grid-cols-2 gap-6 text-sm">
<div class="space-y-3">
<div class="bg-stone-50 rounded-lg p-4">
<p class="text-stone-700">
<strong>[36]</strong>
<a href="https://arxiv.org/abs/2511.19757" class="text-blue-600 hover:underline" target="_blank">
What does it mean to understand language?
</a>
<br/>
<span class="text-stone-500">arXiv preprint</span>
</p>
</div>
<div class="bg-stone-50 rounded-lg p-4">
<p class="text-stone-700">
<strong>[41]</strong>
<a href="https://smallake.kr/wp-content/uploads/2025/12/2511.19757v1.pdf" class="text-blue-600 hover:underline" target="_blank">
Language understanding and situation models
</a>
<br/>
<span class="text-stone-500">Research Paper</span>
</p>
</div>
<div class="bg-stone-50 rounded-lg p-4">
<p class="text-stone-700">
<strong>[45]</strong>
<a href="https://arxiv.org/html/2511.19757v1" class="text-blue-600 hover:underline" target="_blank">
Neural basis of language processing
</a>
<br/>
<span class="text-stone-500">Neuroscience Review</span>
</p>
</div>
<div class="bg-stone-50 rounded-lg p-4">
<p class="text-stone-700">
<strong>[48]</strong>
<a href="https://medium.com/<span class="mention-invalid">@riazleghari</span>/why-ai-still-cant-truly-understand-language-b5931972e4ab" class="text-blue-600 hover:underline" target="_blank">
Why AI Still Can't Truly Understand Language
</a>
<br/>
<span class="text-stone-500">Medium Analysis</span>
</p>
</div>
</div>
<div class="space-y-3">
<div class="bg-stone-50 rounded-lg p-4">
<p class="text-stone-700">
<strong>[58]</strong>
<a href="https://www.pnas.org/doi/10.1073/pnas.2105646118" class="text-blue-600 hover:underline" target="_blank">
Language model training objectives
</a>
<br/>
<span class="text-stone-500">PNAS Journal</span>
</p>
</div>
<div class="bg-stone-50 rounded-lg p-4">
<p class="text-stone-700">
<strong>[63]</strong>
<a href="https://www.tsinghua.edu.cn/info/1182/121813.htm" class="text-blue-600 hover:underline" target="_blank">
清华大学万格智能团队智人HRM模型
</a>
<br/>
<span class="text-stone-500">Tsinghua University Research</span>
</p>
</div>
<div class="bg-stone-50 rounded-lg p-4">
<p class="text-stone-700">
<strong>[37]</strong>
<a href="https://chatpaper.com/zh-CN/paper/213343" class="text-blue-600 hover:underline" target="_blank">
Export Hypothesis Discussion
</a>
<br/>
<span class="text-stone-500">Academic Discussion</span>
</p>
</div>
<div class="bg-stone-50 rounded-lg p-4">
<p class="text-stone-700">
<strong>[38]</strong>
<a href="https://huggingface.co/papers/2511.19757" class="text-blue-600 hover:underline" target="_blank">
Language Understanding Research
</a>
<br/>
<span class="text-stone-500">Hugging Face Papers</span>
</p>
</div>
</div>
</div>
</div>
</div>
<script>
// TOC Toggle and Smooth Scrolling
document.addEventListener('DOMContentLoaded', function() {
const tocToggle = document.getElementById('toc-toggle');
const tocNav = document.getElementById('toc-nav');
const tocLinks = document.querySelectorAll('#toc-nav a');
// TOC Toggle
if (tocToggle && tocNav) {
tocToggle.addEventListener('click', function() {
tocNav.classList.toggle('-translate-x-full');
});
// Close TOC when clicking outside on mobile
document.addEventListener('click', function(event) {
if (window.innerWidth >= 1024) return;
const isClickInsideToc = tocNav.contains(event.target);
const isClickOnToggle = tocToggle.contains(event.target);
if (!isClickInsideToc && !isClickOnToggle && !tocNav.classList.contains('-translate-x-full')) {
tocNav.classList.add('-translate-x-full');
}
});
}
// Smooth Scrolling for TOC Links
tocLinks.forEach(link => {
link.addEventListener('click', function(e) {
e.preventDefault();
const targetId = this.getAttribute('href');
const targetElement = document.querySelector(targetId);
if (targetElement) {
targetElement.scrollIntoView({
behavior: 'smooth',
block: 'start'
});
// Close mobile TOC after clicking
if (window.innerWidth < 1024) {
tocNav.classList.add('-translate-x-full');
}
}
});
});
// Highlight active section in TOC
const sections = document.querySelectorAll('section[id]');
const observerOptions = {
rootMargin: '-20% 0px -70% 0px'
};
const observer = new IntersectionObserver((entries) => {
entries.forEach(entry => {
if (entry.isIntersecting) {
// Remove active class from all links
tocLinks.forEach(link => {
link.classList.remove('text-blue-600', 'bg-blue-50');
link.classList.add('text-stone-600');
});
// Add active class to current link
const activeLink = document.querySelector(`#toc-nav a[href="#${entry.target.id}"]`);
if (activeLink) {
activeLink.classList.remove('text-stone-600');
activeLink.classList.add('text-blue-600', 'bg-blue-50');
}
}
});
}, observerOptions);
sections.forEach(section => observer.observe(section));
// Initialize Mermaid with enhanced theming
mermaid.initialize({
startOnLoad: true,
theme: 'base',
themeVariables: {
primaryColor: '#dbeafe',
primaryTextColor: '#1f2937',
primaryBorderColor: '#374151',
lineColor: '#6b7280',
secondaryColor: '#f3f4f6',
tertiaryColor: '#fef3c7',
background: '#ffffff',
mainBkg: '#ffffff',
secondBkg: '#f9fafb',
tertiaryBkg: '#f3f4f6',
nodeBorder: '#374151',
clusterBkg: '#f3f4f6',
edgeLabelBackground: '#ffffff',
nodeTextColor: '#1f2937',
textColor: '#1f2937',
// Enhanced contrast colors for different node types
cScale0: '#dbeafe',
cScale1: '#f3f4f6',
cScale2: '#fef3c7',
cScale3: '#d1fae5',
cScale4: '#fee2e2',
cScale5: '#e0e7ff',
cScale6: '#f0f9ff',
cScale7: '#fdf2f8',
// Ensure good text contrast on all backgrounds
cScaleLabel0: '#1e3a8a',
cScaleLabel1: '#1f2937',
cScaleLabel2: '#92400e',
cScaleLabel3: '#065f46',
cScaleLabel4: '#991b1b',
cScaleLabel5: '#3730a3',
cScaleLabel6: '#0c4a6e',
cScaleLabel7: '#831843'
},
flowchart: {
useMaxWidth: false,
htmlLabels: true,
curve: 'basis',
padding: 20
},
sequence: {
useMaxWidth: false,
wrap: true
},
gantt: {
useMaxWidth: false
}
});
// Initialize Mermaid Controls for zoom and pan
initializeMermaidControls();
});
// Initialize Mermaid Controls for zoom and pan
function initializeMermaidControls() {
const containers = document.querySelectorAll('.mermaid-container');
containers.forEach(container => {
const mermaidElement = container.querySelector('.mermaid');
let scale = 1;
let isDragging = false;
let startX, startY, translateX = 0, translateY = 0;
// 触摸相关状态
let isTouch = false;
let touchStartTime = 0;
let initialDistance = 0;
let initialScale = 1;
let isPinching = false;
// Zoom controls
const zoomInBtn = container.querySelector('.zoom-in');
const zoomOutBtn = container.querySelector('.zoom-out');
const resetBtn = container.querySelector('.reset-zoom');
const fullscreenBtn = container.querySelector('.fullscreen');
function updateTransform() {
mermaidElement.style.transform = `translate(${translateX}px, ${translateY}px) scale(${scale})`;
if (scale > 1) {
container.classList.add('zoomed');
} else {
container.classList.remove('zoomed');
}
mermaidElement.style.cursor = isDragging ? 'grabbing' : 'grab';
}
if (zoomInBtn) {
zoomInBtn.addEventListener('click', () => {
scale = Math.min(scale * 1.25, 4);
updateTransform();
});
}
if (zoomOutBtn) {
zoomOutBtn.addEventListener('click', () => {
scale = Math.max(scale / 1.25, 0.3);
if (scale <= 1) {
translateX = 0;
translateY = 0;
}
updateTransform();
});
}
if (resetBtn) {
resetBtn.addEventListener('click', () => {
scale = 1;
translateX = 0;
translateY = 0;
updateTransform();
});
}
if (fullscreenBtn) {
fullscreenBtn.addEventListener('click', () => {
if (container.requestFullscreen) {
container.requestFullscreen();
} else if (container.webkitRequestFullscreen) {
container.webkitRequestFullscreen();
} else if (container.msRequestFullscreen) {
container.msRequestFullscreen();
}
});
}
// Mouse Events
mermaidElement.addEventListener('mousedown', (e) => {
if (isTouch) return; // 如果是触摸设备，忽略鼠标事件
isDragging = true;
startX = e.clientX - translateX;
startY = e.clientY - translateY;
mermaidElement.style.cursor = 'grabbing';
updateTransform();
e.preventDefault();
});
document.addEventListener('mousemove', (e) => {
if (isDragging && !isTouch) {
translateX = e.clientX - startX;
translateY = e.clientY - startY;
updateTransform();
}
});
document.addEventListener('mouseup', () => {
if (isDragging && !isTouch) {
isDragging = false;
mermaidElement.style.cursor = 'grab';
updateTransform();
}
});
document.addEventListener('mouseleave', () => {
if (isDragging && !isTouch) {
isDragging = false;
mermaidElement.style.cursor = 'grab';
updateTransform();
}
});
// 获取两点之间的距离
function getTouchDistance(touch1, touch2) {
return Math.hypot(
touch2.clientX - touch1.clientX,
touch2.clientY - touch1.clientY
);
}
// Touch Events - 触摸事件处理
mermaidElement.addEventListener('touchstart', (e) => {
isTouch = true;
touchStartTime = Date.now();
if (e.touches.length === 1) {
// 单指拖动
isPinching = false;
isDragging = true;
const touch = e.touches[0];
startX = touch.clientX - translateX;
startY = touch.clientY - translateY;
} else if (e.touches.length === 2) {
// 双指缩放
isPinching = true;
isDragging = false;
const touch1 = e.touches[0];
const touch2 = e.touches[1];
initialDistance = getTouchDistance(touch1, touch2);
initialScale = scale;
}
e.preventDefault();
}, { passive: false });
mermaidElement.addEventListener('touchmove', (e) => {
if (e.touches.length === 1 && isDragging && !isPinching) {
// 单指拖动
const touch = e.touches[0];
translateX = touch.clientX - startX;
translateY = touch.clientY - startY;
updateTransform();
} else if (e.touches.length === 2 && isPinching) {
// 双指缩放
const touch1 = e.touches[0];
const touch2 = e.touches[1];
const currentDistance = getTouchDistance(touch1, touch2);
if (initialDistance > 0) {
const newScale = Math.min(Math.max(
initialScale * (currentDistance / initialDistance),
0.3
), 4);
scale = newScale;
updateTransform();
}
}
e.preventDefault();
}, { passive: false });
mermaidElement.addEventListener('touchend', (e) => {
// 重置状态
if (e.touches.length === 0) {
isDragging = false;
isPinching = false;
initialDistance = 0;
// 延迟重置isTouch，避免鼠标事件立即触发
setTimeout(() => {
isTouch = false;
}, 100);
} else if (e.touches.length === 1 && isPinching) {
// 从双指变为单指，切换为拖动模式
isPinching = false;
isDragging = true;
const touch = e.touches[0];
startX = touch.clientX - translateX;
startY = touch.clientY - translateY;
}
updateTransform();
});
mermaidElement.addEventListener('touchcancel', (e) => {
isDragging = false;
isPinching = false;
initialDistance = 0;
setTimeout(() => {
isTouch = false;
}, 100);
updateTransform();
});
// Enhanced wheel zoom with better center point handling
container.addEventListener('wheel', (e) => {
e.preventDefault();
const rect = container.getBoundingClientRect();
const centerX = rect.width / 2;
const centerY = rect.height / 2;
const delta = e.deltaY > 0 ? 0.9 : 1.1;
const newScale = Math.min(Math.max(scale * delta, 0.3), 4);
// Adjust translation to zoom towards center
if (newScale !== scale) {
const scaleDiff = newScale / scale;
translateX = translateX * scaleDiff;
translateY = translateY * scaleDiff;
scale = newScale;
if (scale <= 1) {
translateX = 0;
translateY = 0;
}
updateTransform();
}
});
// Initialize display
updateTransform();
});
}
</script>
</body></html>
登录后可参与表态