一个为所有对 AI 发展感兴趣的人打造的 AI / LLM / RAG / Agent 专业名词百科，尝试用不吓人的语言，帮你看懂「飞速发展的 AI 世界到底在做什么」。
既是你查概念时的「AI 术语字典」，也是你梳理知识脉络的「导航页」——无论你是刚入坑的学习者，还是一线实践者，都可以在这里对齐「我们到底在说什么」。
在 AI 领域，名词爆炸式增长：从基础模型、训练范式，到 RAG、Agent、多模态工作流，很容易「听过很多词，但记不清含义和关系」。
这个项目想做的是一件简单但不容易的事：
- 把跑得很快的前沿名词，翻译成尽量清晰、克制的中文说明；
- 按主题和场景整理，帮你建立一张「AI 概念地图」；
- 通过社区协作，持续更新、修正、补充，让术语跟得上实践；
- 借助 GitHub 账号和内置评论区，让对同一术语感兴趣的人在同一页下相遇、讨论，并顺手互相「点进主页看看对方在做什么」。
它不是「权威标准」，而是一份开放的公共笔记本：任何人都可以提出不同理解、补充案例或更好的说法，一起把它打磨得更好懂；评论区里留下的交流和争论，也会慢慢成为这份笔记的一部分。
你可以：
- 在首页通过「搜索 / 分类 / 视图切换」快速浏览；
- 点击卡片进入术语详情，查看更完整解释与外链；
- 在详情页使用 giscus 发表评论、补充理解。
作为读者 / 学习者：
- 在首页搜索你关心的术语（中英文、别名或关键词均可）；
- 使用分类筛选探索某一条线，例如
foundation、rag、agent等； - 点击任一术语卡片，进入详情页阅读更完整的解释与外链；
- 在术语页底部使用评论区，分享你的理解、补充例子，或指出你觉得哪里没讲清楚。
作为团队 / 组织：
- 把本仓库视为「上游术语库」，让团队成员在这里统一提 PR 共建，方便知识持续沉淀与更新；
- 在你们内部文档、知识库或培训材料中引用本百科的链接，从同一套术语出发沟通，减少「同词不同意 / 同意不同词」；
- 如需更改站点外观或部署方式，可参考 GitHub Pages / Jekyll 文档，自行按需配置。
欢迎你：
- 补充新的术语或改进已有解释；
- 调整分类与关键词，使索引更好用；
- 改进样式与交互，让阅读体验更顺滑。 无论你是第一次用 GitHub，还是已经习惯各种工程协作，只要你愿意让某个概念更好懂，这里都非常欢迎。
推荐的贡献方式（零门槛版）：
- 在站点上点击「在线编辑」/「编辑此词条」按钮，或在 GitHub 的
_terms/目录中直接编辑/新建 Markdown 文件； - 按模板填写或修改内容，提交时简单写几句你做了什么调整；
- GitHub 会引导你完成变更提案（Pull Request）的创建，我们会在 Review 后合并到主站。
所有改动都会集中到这一份公共百科中，方便长期维护与演进，也方便后来者直接受益于你的贡献。
如果你想了解术语文件的 Front Matter 模板、仓库结构、本地开发与 giscus 配置等更细节的内容，请查看 DEV.md。
面向进阶贡献者和维护者的更完整说明（例如：项目结构、术语写作模板、本地预览命令、前端与评论系统配置等），请阅读 DEV.md。
AI Glossary is an open, community-driven knowledge base for AI / LLM / RAG / Agent terminology. It is for anyone curious about how AI is evolving—not just engineers—and aims to explain “what on earth is actually happening” in this fast-moving space.
You can treat it both as a “reference dictionary” when you run into an unfamiliar term and as a “navigation page” when you are trying to build a bigger-picture mental model.
The AI landscape churns out new buzzwords daily: model architectures, training strategies, RAG pipelines, agents, multimodal workflows, and beyond. It is very easy to feel like you have “heard the word before” but cannot recall what it really means or how it connects to others.
This glossary tries to do a few simple things well:
- Offer clear, down-to-earth explanations instead of marketing speak;
- Organize terms by topic and use case so you can build a mental map;
- Continuously evolve through community contributions as practices change.
It is not meant to be a “final authority”, but an open notebook where people can bring in their understanding, references, and counterpoints—and gradually make the concepts easier to learn.
- Live site: https://notebookai.github.io/ai-wiki/
You can:
- Browse through full-text search, category filters, or card/list view toggles;
- Dive into a term page for detailed explanations and references;
- Leave comments via giscus at the bottom of each term page.
Readers / Learners
- Search any keyword, alias, or English/Chinese name on the home page;
- Filter by categories like
foundation,rag,agent, etc.; - Open a card to read the full entry, examples, and related resources;
- Use the giscus comments to share your understanding, ask questions, or point out where the explanation could be clearer.
Teams / Organizations
- Treat this repo as the upstream “shared glossary” so your team can open PRs here and keep terminology aligned in one place;
- Reference this glossary from your internal docs, wikis, or training material so everyone talks about the same vocabulary;
- If you need a customized deployment or appearance, refer to GitHub Pages / Jekyll documentation and configure things as needed.
We welcome:
- New or improved term entries;
- Better taxonomy / keyword coverage;
- Styling or UX improvements.
Whether you are new to GitHub or already used to engineering workflows, as long as you want to help make concepts easier to understand, you are very welcome here.
Recommended contribution flow (no heavy setup required):
- Use the “Edit this term” / “Add term” buttons on the site, or edit/create Markdown files under
_terms/via GitHub’s web UI; - Add or adjust content based on the term template, and include a short note describing your changes when you submit;
- GitHub will walk you through creating a Pull Request, and we will review and merge it into the main site.
All improvements land in this single upstream glossary, which makes it easier to maintain and evolve over time and lets future readers benefit directly from your work.
For details about term templates, repository structure, local development, and giscus configuration, please refer to DEV.md.
If you want to dive deeper into how this site is structured, how term files are organized, or how to run and tweak the project locally, check out DEV.md.
