Eagle View · 鹰瞰大图
An eagle-eye view of LLM Primer's full knowledge structure. Hover a module card to highlight learning paths; click any article link to jump directly in.
LLM Primer 全站结构鹰瞰。悬停模块卡片可高亮学习路径，点击文章链接直达正文。
- Tokenization: How Models See Text
- Attention: Choosing the Relevant Context
- Sampling and Decoding: From Probabilities to Text
- Transformer Architecture: The Skeleton of Modern LLMs
- Embeddings: Putting Discrete Symbols into Continuous Space
- Positional Encoding: Where Does Order Come From
- Why LLMs Emerge Abilities
- RAG and Retrieval Augmentation: Giving Models External Memory
- Agents and Tool Use: Models Are More Than Chat
- Prompt Engineering: The Art of Talking to Models
- Evaluation and Benchmarks: Judging Model Quality
- Safety and Adversarial: Protecting and Attacking Models
- Code Generation: How Models Write Programs
── Recommended learning path ╌╌ Optional cross-module link