Prompt Engineering Techniques: LLM 优化指南

Prompt engineering 是通过系统化设计输入来获得 AI 模型期望输出的艺术与科学。正确的提示策略可将模型性能提升高达 50%。

Prompt Engineering 基础

Prompt 结构

一个有效的 Prompt 通常包含以下组件：

System Instruction：模型角色与行为规则
Context：与任务相关的背景信息
Example(s)：期望输出格式示例
Task：具体请求或问题
Output Format：期望的响应结构

基础 Prompt 模板

1[Role Definition]
2You are a {field of expertise} expert.
3
4[Context]
5{Relevant information}
6
7[Task]
8{Specific request}
9
10[Format]
11{Expected output format}

基础 Prompt 技巧

1. Zero-Shot Prompting（零样本提示）

在不提供示例的情况下直接定义任务：

Summarize the following text:
{text}

适用场景：

简单任务
常识性问题
分类任务

2. Few-Shot Prompting（少样本提示）

通过示例来定义任务：

1Text: "This product is great, I am very satisfied"
2Sentiment: Positive
3
4Text: "It was a terrible experience, I do not recommend it"
5Sentiment: Negative
6
7Text: "It's okay for the price"
8Sentiment: ?

最佳实践：

通常 3-5 个示例即可
覆盖各种边界情况
示例顺序随机化

3. Chain-of-Thought (CoT)

逐步思考：

1Question: A store has 15 apples. 8 apples were sold, 
2then 6 more apples arrived. How many apples are there?
3
4Let's think step by step:
51. Initially there are 15 apples
62. 8 apples sold: 15 - 8 = 7 apples left
73. 6 apples arrived: 7 + 6 = 13 apples
8
9Answer: 13 apples

4. Self-Consistency

为同一问题生成不同推理路径：

1Solve this problem in 3 different ways and choose the most consistent answer:
2
3[Problem]

5. Tree of Thoughts (ToT)

分支式思维树：

1Problem: {complex problem}
2
3Thought 1: {approach A}
4  → Sub-thought 1.1: {detail}
5  → Sub-thought 1.2: {detail}
6  
7Thought 2: {approach B}
8  → Sub-thought 2.1: {detail}
9  
10Evaluate and select the most suitable path.

高级技巧

ReAct（Reasoning + Acting）

思考与行动循环：

1Question: How many times larger is Istanbul's population than Paris?
2
3Thought: I need to find the population of both cities
4Action: [search] Istanbul population
5Observation: Istanbul population ~16 million
6
7Thought: Now I need to find Paris population
8Action: [search] Paris population  
9Observation: Paris population ~2.2 million
10
11Thought: I can calculate the ratio
12Action: [calculate] 16 / 2.2
13Observation: 7.27
14
15Answer: Istanbul's population is approximately 7.3 times larger than Paris.

Constitutional AI Prompting

定义伦理与安全规则：

1System: You are a helpful assistant.
2
3Rules:
41. Do not generate harmful content
52. Do not share personal information
63. Do not help with illegal activities
74. Always be honest
8
9User question: {question}

Role Prompting

定义专业领域角色：

1You are a cybersecurity expert with 20 years of experience.
2You have worked as a CISO in Fortune 500 companies.
3You can explain technical details clearly and understandably.
4
5User's question: {question}
6## 提示优化策略
7
8### 1. 提高具体性
9
10❌ 不佳示例：

Write a blog post


✅ 良好示例：

Target audience: Software developers Topic: Docker container security Length: 1500-2000 words Tone: Technical but accessible Format: Introduction, 5 main sections, conclusion

1
2### 2. 确定输出格式
3

Provide your response in this JSON format: { "summary": "string", "key_points": ["string"], "next_steps": ["string"], "confidence_score": number }

1
2### 3. 负向提示
3
4指定不希望出现的行为：
5

Do NOT do the following:

Give speculative information
Make claims without citing sources
Lead the user
Give excessively long answers

1
2### 4. 使用分隔符
3
4明确不同部分：
5

###CONTEXT### {context information}

###TASK### {work to be done}

###FORMAT### {output format}

1
2## 模型相关优化
3
4### 针对 GPT
5

Use System message effectively
Activate JSON mode: response_format={"type": "json_object"}
Temperature: 0.7-1.0 for creative tasks, 0.1-0.3 for analytical

1
2### 针对 Claude
3

Use XML tags: <context>, <task>, <format>
Put important information at the end in long context
Evaluate Thinking tags

1
2### 针对 Gemini
3

Optimize for multimodal prompts
Up-to-date information with Grounding
Adjust Safety settings

1
2## 提示测试与迭代
3
4### A/B 测试框架
5

Prompt A: {version 1} Prompt B: {version 2}

Metrics:

Accuracy: %
Consistency: 1-5
Latency: ms
Token usage: #

1
2### 提示版本管理
3

prompt_v1.0: First version prompt_v1.1: Typo corrections prompt_v2.0: CoT added prompt_v2.1: Output format changed

1
2## 企业级提示管理
3
4### 创建提示库
5

/prompts /classification - sentiment_analysis.json - intent_detection.json /generation - blog_writer.json - code_reviewer.json /extraction - entity_extraction.json - data_parsing.json

1
2### 提示模板系统
3
4```python
5class PromptTemplate:
6    def __init__(self, template, variables):
7        self.template = template
8        self.variables = variables
9    
10    def render(self, **kwargs):
11        return self.template.format(**kwargs)
12
13# Usage
14sentiment_prompt = PromptTemplate(
15    template="Analyze sentiment: {text}",
16    variables=["text"]
17)

常见错误与解决方案

错误 1：提示过于模糊

问题： 模型无法理解你的需求
解决方案： 添加明确的指令和示例

错误 2：提示过长

问题： 超出 Token 限制，成本升高
解决方案： 删除不必要信息，使用摘要

错误 3：指令冲突

问题： 模型行为不一致
解决方案： 优先级排序并澄清规则

错误 4：幻觉

问题： 模型生成不存在的信息
解决方案： 使用 Grounding、要求引用、降低 temperature

总结

提示工程是 AI 项目成功的关键因素之一。通过正确的技术和系统化的方法，你可以显著提升模型性能。

在 Veni AI，我们为企业客户构建定制化提示策略。欢迎联系我们获取专业支持。

提示工程技术：LLM 优化指南