如何确保大模型 RAG 生成的信息是基于可靠的数据源？

news2025/5/25 6:14:01

在不断发展的人工智能 (AI) 领域中，检索增强生成 (RAG) 已成为一种强大的技术。

RAG 弥合了大型语言模型 (LLM) 与外部知识源之间的差距，使 AI 系统能够提供更全面和信息丰富的响应。然而，一个关键因素有时会缺失——透明性。

我们如何能够确定 RAG 系统呈现的信息是基于可靠来源的？

本文介绍了一种引人注目的解决方案：使用结构化生成的带源突出显示的 RAG。这种创新的方法不仅利用了 RAG 检索相关信息的能力，还突出了支持生成答案的具体来源。喜欢本文记得收藏、点赞、关注，希望大模型技术交流的文末加入我们。

理解基础构件

在深入探讨之前，让我们先建立核心概念的基础：

结构化生成：这种技术引导大型语言模型 (LLM) 的输出遵循预定义的结构。可以将其想象为为 LLM 提供一张路线图，确保生成的文本符合特定格式。

带源突出显示的 RAG 的优势

RAG 和结构化生成的整合提供了诸多优点：

增强信任和透明性：突出显示的来源使用户能够评估所呈现信息的可信度。这有助于培养对系统的信任，并使用户能够深入了解支持证据。
改善可解释性：通过明确指出答案背后的来源，系统变得更加透明。用户能够深入了解推理过程，便于调试和进一步探索知识库。
更广泛的适用性：这种方法适用于用户不仅需要答案，还需要理由和清晰的审计记录的场景。它在教育、研究和法律领域尤为有价值。

代码实现

让我们深入了解使用结构化生成的带源突出显示的 RAG。

步骤 I：安装库

!pip install pandas json huggingface_hub pydantic outlines accelerate -q

步骤 II：导入库

import pandas as pd
import json
from huggingface_hub import InferenceClient

pd.set_option("display.max_colwidth", None)

repo_id = "meta-llama/Meta-Llama-3-8B-Instruct"

llm_client = InferenceClient(model=repo_id, timeout=120)

步骤 III：提示模型

RELEVANT_CONTEXT = """
Document:

The weather is really nice in Paris today.
To define a stop sequence in Transformers, you should pass the stop_sequence argument in your pipeline or model.
"""

RAG_PROMPT_TEMPLATE_JSON = """
Answer the user query based on the source documents.

Here are the source documents: {context}

You should provide your answer as a JSON blob, and also provide all relevant short source snippets from the documents on which you directly based your answer, and a confidence score as a float between 0 and 1.
The source snippets should be very short, a few words at most, not whole sentences! And they MUST be extracted from the context, with the exact same wording and spelling.

Your answer should be built as follows, it must contain the "Answer:" and "End of answer." sequences.

Answer:
{{
  "answer": your_answer,
  "confidence_score": your_confidence_score,
  "source_snippets": ["snippet_1", "snippet_2", ...]
}}
End of answer.

Now begin!
Here is the user question: {user_query}.
Answer:
"""

USER_QUERY = "How can I define a stop sequence in Transformers?" 

prompt = RAG_PROMPT_TEMPLATE_JSON.format(context=RELEVANT_CONTEXT, user_query=USER_QUERY)

print(prompt)

输出：

Answer the user query based on the source documents.

Here are the source documents: 
Document:

The weather is really nice in Paris today.
To define a stop sequence in Transformers, you should pass the stop_sequence argument in your pipeline or model.

You should provide your answer as a JSON blob, and also provide all relevant short source snippets from the documents on which you directly based your answer, and a confidence score as a float between 0 and 1.
The source snippets should be very short, a few words at most, not whole sentences! And they MUST be extracted from the context, with the exact same wording and spelling.

Your answer should be built as follows, it must contain the "Answer:" and "End of answer." sequences.

Answer:
{
  "answer": your_answer,
  "confidence_score": your_confidence_score,
  "source_snippets": ["snippet_1", "snippet_2", ...]
}
End of answer.

Now begin!
Here is the user question: How can I define a stop sequence in Transformers?.
Answer:

继续代码：

answer = llm_client.text_generation(
    prompt,
    max_new_tokens=1000,
)

answer = answer.split("End of answer.")[0]
print(answer)

输出：

{
  "answer": "You should pass the stop_sequence argument in your pipeline or model.",
  "confidence_score": 0.9,
  "source_snippets": ["stop_sequence", "pipeline or model"]
}

步骤 IV：受限解码

from pydantic import BaseModel, confloat, StringConstraints
from typing import List, Annotated

class AnswerWithSnippets(BaseModel):
    answer: Annotated[str, StringConstraints(min_length=10, max_length=100)]
    confidence: Annotated[float, confloat(ge=0.0, le=1.0)]
    source_snippets: List[Annotated[str, StringConstraints(max_length=30)]]

# Using text_generation
answer = llm_client.text_generation(
    prompt,
    grammar={"type": "json", "value": AnswerWithSnippets.schema()},
    max_new_tokens=250,
    temperature=1.6,
    return_full_text=False,
)
print(answer)

# Using post
data = {
    "inputs": prompt,
    "parameters": {
        "temperature": 1.6,
        "return_full_text": False,
        "grammar": {"type": "json", "value": AnswerWithSnippets.schema()},
        "max_new_tokens": 250,
    },
}
answer = json.loads(llm_client.post(json=data))[0]["generated_text"]
print(answer)

输出：

{
  "answer": "You should pass the stop_sequence argument in your modemÏallerbate hassceneable measles updatedAt原因",
  "confidence": 0.9,
  "source_snippets": ["in Transformers", "stop_sequence argument in your"]
}

{
"answer": "To define a stop sequence in Transformers, you should pass the stop-sequence argument in your...giÃ",  
"confidence": 1,  
"source_snippets": ["seq이야","stration nhiên thị ji是什么hpeldo"]
}