2.1 概览
本章主要总结了 LLM Prompt Engineering 相关的技术、原则、模版 和 工具。
参考资料：
- Prompt Engineering Guide
- Learn Prompting
- Prompt Engineering Institute
1. Techniques
| Technique | description |
|---|
| Zero-Shot Prompting | |
| Few-Shot Prompting | |
| Chain-of-Thought (COT) Prompting | |
| Tree of Thoughts (ToT) Prompting | |
| Generated Knowledge Prompting | |
| {.table .table-bordered .border-secondary} | |
2. Principles
| Principle | description |
|---|
| start simple and gradually interative | |
| Write clear and specific instructions | |
| Give the model time to “think” | |
| {.table .table-bordered .border-secondary} | |
3. Templates
3.1 Basic Template
| components | description | nessesary? |
|---|
| Instruction | a specific task or instruction you want the model to perform. | Yes |
| Context | external information or additional context that can steer the model to better responses. | Optional |
| Input | the input or question that we are interested to find a response for | Optional |
| Outcome | the type or format of the output. | Optional |
| {.table .table-bordered .border-secondary} | | |
3.2 Advanced Template
| components | description | nessesary? |
|---|
| Role | Specify ChatGPT’s role. | Yes |
| Instruction | a specific task or instruction you want the model to perform | Yes |
| Context | external information or additional context that can steer the model to better responses. | Optional |
| Example | give examples to illustrate the input and outcome, or examples of probolem solving process. | Optional |
| Input | the input or question that we are interested to find a response for | Optional |
| Outcome | the type, format or length of the output. | Optional |
| {.table .table-bordered .border-secondary} | | |
4. Tools