-
Notifications
You must be signed in to change notification settings - Fork 1
Expand file tree
/
Copy pathappendix-b.html
More file actions
342 lines (327 loc) · 12 KB
/
Copy pathappendix-b.html
File metadata and controls
342 lines (327 loc) · 12 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
<!DOCTYPE html>
<html lang="zh-CN">
<head>
<meta charset="UTF-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<meta name="description" content="AI学习笔记 - 从零开始系统学习人工智能，涵盖前置基础、核心原理、工程实践、应用工具、进阶展望">
<meta name="keywords" content="AI, 人工智能, 机器学习, 深度学习, Python, 数学基础">
<meta name="author" content="AI Learner">
<title>附录B：术语表 - AI学习笔记</title>
<!-- Google Fonts - Inter -->
<link rel="preconnect" href="https://fonts.googleapis.com">
<link rel="preconnect" href="https://fonts.gstatic.com">
<link href="https://fonts.googleapis.com/css2?family=Inter:wght@300;400;500;600;700;800&display=swap" rel="stylesheet">
<!-- Font Awesome Icons -->
<link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.5.1/css/all.min.css">
<!-- KaTeX - 轻量级数学公式渲染 -->
<link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/katex@0.16.9/dist/katex.min.css">
<script defer src="https://cdn.jsdelivr.net/npm/katex@0.16.9/dist/katex.min.js"></script>
<script defer src="https://cdn.jsdelivr.net/npm/katex@0.16.9/dist/contrib/auto-render.min.js"
onload="renderMathInElement(document.body, {
delimiters: [
{left: '$$', right: '$$', display: true},
{left: '\\[', right: '\\]', display: true},
{left: '\\(', right: '\\)', display: false},
{left: '$', right: '$', display: false}
],
throwOnError: false
});"></script>
<!-- Prism.js 代码高亮 - tomorrow-night 主题 -->
<link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/prism/1.29.0/themes/prism-tomorrow.min.css">
<script src="https://cdnjs.cloudflare.com/ajax/libs/prism/1.29.0/prism.min.js"></script>
<script src="https://cdnjs.cloudflare.com/ajax/libs/prism/1.29.0/components/prism-python.min.js"></script>
<script src="https://cdnjs.cloudflare.com/ajax/libs/prism/1.29.0/components/prism-bash.min.js"></script>
<script src="https://cdnjs.cloudflare.com/ajax/libs/prism/1.29.0/plugins/copy-to-clipboard/prism-copy-to-clipboard.min.js"></script>
<!-- 自定义样式表 -->
<link rel="stylesheet" href="../../css/main.css">
<link rel="stylesheet" href="../../css/dark-mode.css">
<link rel="stylesheet" href="../../css/mobile-optimizations.css">
<!-- 内联CSS样式 -->
<!-- AI-Notes共享样式 -->
<link rel="stylesheet" href="../assets/ai-notes.css">
</head>
<body>
<!-- ==================== 导航栏 ==================== -->
<nav class="ai-nav">
<a href="../../index.html" class="ai-nav-brand">
<i class="fas fa-brain"></i>
AI学习笔记
</a>
<ul class="ai-nav-links" id="navLinks">
<li><a href="../../index.html"><i class="fas fa-home"></i> 首页</a></li>
<li><a href="ch15.html"><i class="fas fa-robot"></i> Agent</a></li>
<li><a href="ch16.html"><i class="fas fa-tools"></i> 工具</a></li>
<li><a href="ch17.html"><i class="fas fa-book-open"></i> 知识库</a></li>
<li><a href="ch18.html"><i class="fas fa-layer-group"></i> 记忆</a></li>
<li><a href="ch19.html"><i class="fas fa-comment-dots"></i> 提示</a></li>
<li><a href="ch20.html"><i class="fas fa-database"></i> 数据</a></li>
<li><a href="ch21.html"><i class="fas fa-exclamation-circle"></i> 局限</a></li>
<li><a href="ch22.html"><i class="fas fa-toolbox"></i> 平台</a></li>
<li><a href="ch23.html"><i class="fas fa-history"></i> 简史</a></li>
<li><a href="ch24.html"><i class="fas fa-rocket"></i> 实战</a></li>
</ul>
<button class="ai-nav-toggle" id="navToggle" aria-label="切换导航">
<i class="fas fa-bars"></i>
</button>
</nav>
<!-- ==================== Hero Section ==================== -->
<section class="ai-hero">
<h1><i class="fas fa-brain"></i> AI 学习笔记</h1>
<p>附录B：术语表</p>
<div class="ai-hero-badges">
<span class="ai-hero-badge"><i class="fas fa-book"></i> 系统学习</span>
<span class="ai-hero-badge"><i class="fas fa-code"></i> 实践导向</span>
</div>
</section>
<!-- ==================== 主内容区域 ==================== -->
<div class="content-wrapper">
<!-- ========== 主内容 ========== -->
<main class="main-content">
<nav class="ai-chapter-nav">
<div class="breadcrumb">
<a href="../../index.html"><i class="fas fa-home"></i> 首页</a>
<span class="separator">/</span>
<span class="current">附录B：术语表</span>
</div>
<div class="chapter-nav-links">
<a href="appendix-a.html" class="prev-chapter"><i class="fas fa-arrow-left"></i> 上一章</a>
<a href="ch24.html" class="next-chapter">下一章 <i class="fas fa-arrow-right"></i></a>
</div>
</nav>
<!-- 附录B：术语表 -->
<!-- ============================================================ -->
<h2 class="ai-section-title"><i class="fas fa-book"></i> 附录B：术语表</h2>
<div class="concept-section">
<table class="comparison-table">
<tr>
<th>术语</th>
<th>英文</th>
<th>简短解释</th>
</tr>
<tr>
<td>人工智能</td>
<td>Artificial Intelligence (AI)</td>
<td>让计算机模拟人类智能行为的技术</td>
</tr>
<tr>
<td>机器学习</td>
<td>Machine Learning (ML)</td>
<td>从数据中自动学习规律和模式的算法</td>
</tr>
<tr>
<td>深度学习</td>
<td>Deep Learning (DL)</td>
<td>使用多层神经网络的学习方法</td>
</tr>
<tr>
<td>大语言模型</td>
<td>Large Language Model (LLM)</td>
<td>参数量巨大的预训练语言模型</td>
</tr>
<tr>
<td>Transformer</td>
<td>Transformer</td>
<td>基于自注意力机制的神经网络架构</td>
</tr>
<tr>
<td>注意力机制</td>
<td>Attention Mechanism</td>
<td>让模型关注输入中重要部分的技术</td>
</tr>
<tr>
<td>Token</td>
<td>Token</td>
<td>文本处理的基本单元，可以是一个字、词或子词</td>
</tr>
<tr>
<td>Embedding</td>
<td>Embedding</td>
<td>将非数值数据（如文字）转换为数值向量的过程</td>
</tr>
<tr>
<td>向量数据库</td>
<td>Vector Database</td>
<td>专门存储和检索高维向量的数据库</td>
</tr>
<tr>
<td>余弦相似度</td>
<td>Cosine Similarity</td>
<td>衡量两个向量方向相似性的指标</td>
</tr>
<tr>
<td>RAG</td>
<td>Retrieval-Augmented Generation</td>
<td>检索增强生成，结合信息检索和文本生成</td>
</tr>
<tr>
<td>AI智能体</td>
<td>AI Agent</td>
<td>能自主感知、规划、执行和记忆的AI系统</td>
</tr>
<tr>
<td>ReAct</td>
<td>Reasoning + Acting</td>
<td>交替进行推理和行动的Agent模式</td>
</tr>
<tr>
<td>Function Calling</td>
<td>Function Calling</td>
<td>让LLM调用外部函数的机制</td>
</tr>
<tr>
<td>思维链</td>
<td>Chain of Thought (CoT)</td>
<td>引导LLM逐步推理的提示技术</td>
</tr>
<tr>
<td>上下文窗口</td>
<td>Context Window</td>
<td>LLM一次能处理的最大Token数量</td>
</tr>
<tr>
<td>幻觉</td>
<td>Hallucination</td>
<td>AI生成看似合理但实际不正确的内容</td>
</tr>
<tr>
<td>对齐</td>
<td>Alignment</td>
<td>确保AI行为符合人类价值观的过程</td>
</tr>
<tr>
<td>RLHF</td>
<td>Reinforcement Learning from Human Feedback</td>
<td>基于人类反馈的强化学习，用于AI对齐</td>
</tr>
<tr>
<td>量化</td>
<td>Quantization</td>
<td>将模型参数从高精度转为低精度以减小模型</td>
</tr>
<tr>
<td>剪枝</td>
<td>Pruning</td>
<td>移除模型中不重要的参数或神经元</td>
</tr>
<tr>
<td>知识蒸馏</td>
<td>Knowledge Distillation</td>
<td>将大模型知识转移到小模型的技术</td>
</tr>
<tr>
<td>KV Cache</td>
<td>Key-Value Cache</td>
<td>缓存注意力计算结果以加速推理</td>
</tr>
<tr>
<td>Flash Attention</td>
<td>Flash Attention</td>
<td>优化GPU内存访问加速注意力计算的算法</td>
</tr>
<tr>
<td>多模态</td>
<td>Multimodal</td>
<td>能同时处理文本、图像、音频等多种数据类型</td>
</tr>
<tr>
<td>数据增强</td>
<td>Data Augmentation</td>
<td>通过变换扩充数据集以提升模型泛化能力</td>
</tr>
<tr>
<td>过拟合</td>
<td>Overfitting</td>
<td>模型在训练数据上表现好但在新数据上表现差</td>
</tr>
<tr>
<td>泛化能力</td>
<td>Generalization</td>
<td>模型在面对未见过的数据时的表现能力</td>
</tr>
<tr>
<td>微调</td>
<td>Fine-tuning</td>
<td>在预训练模型基础上用特定数据进一步训练</td>
</tr>
<tr>
<td>LoRA</td>
<td>Low-Rank Adaptation</td>
<td>低秩适配，一种高效的模型微调方法</td>
</tr>
<tr>
<td>显存</td>
<td>VRAM (Video RAM)</td>
<td>GPU上用于存储模型和计算数据的内存</td>
</tr>
<tr>
<td>算力</td>
<td>Computing Power / FLOPS</td>
<td>计算机进行数值计算的能力</td>
</tr>
<tr>
<td>BPE</td>
<td>Byte Pair Encoding</td>
<td>一种子词分词算法，GPT系列使用</td>
</tr>
</table>
</div>
<!-- ============================================================ -->
</main>
</div>
<!-- 页脚 -->
<!-- ============================================================ -->
<footer class="footer">
<div class="footer-content">
<div class="footer-section">
<h3><i class="fas fa-graduation-cap"></i> AI学习笔记</h3>
<p>从零开始系统学习人工智能</p>
<p>涵盖前置基础、核心原理、工程实践、应用工具、进阶展望</p>
</div>
<div class="footer-section">
<h3><i class="fas fa-bookmark"></i> 内容导航</h3>
<p>第1-11章：基础与核心（见 ai-notes-core.html）</p>
<p>第12-23章：工程实践与进阶（本文件）</p>
<p>附录A：GPU与算力指南</p>
<p>附录B：术语表</p>
</div>
<div class="footer-section">
<h3><i class="fas fa-info-circle"></i> 说明</h3>
<p>本笔记仅供学习参考，内容如有错误欢迎指正。</p>
<p>AI技术发展迅速，部分信息可能随时间变化。</p>
</div>
</div>
<div class="footer-bottom">
<p>AI学习笔记 © 2024 | 持续更新中</p>
</div>
</footer>
</div><!-- /ai-notes-container -->
<!-- Scripts -->
<script src="../../js/mobile-menu.js"></script>
<script src="../../js/dark-mode.js"></script>
<script>
// 目录滚动高亮
(function() {
const tocLinks = document.querySelectorAll('.toc-sidebar a');
const sections = [];
tocLinks.forEach(link => {
const id = link.getAttribute('href');
if (id && id.startsWith('#')) {
const el = document.querySelector(id);
if (el) sections.push({ link, el });
}
});
function updateActive() {
const scrollY = window.scrollY + 100;
let current = sections[0];
for (const s of sections) {
if (s.el.offsetTop <= scrollY) current = s;
}
tocLinks.forEach(l => l.classList.remove('active'));
if (current) current.link.classList.add('active');
}
window.addEventListener('scroll', updateActive);
updateActive();
})();
</script>
</body>
</html>