复核引用

通过与源文档核对，捕捉错误或凭空捏造的引用。一个 TypeSafe Choice 问题判断引文的上下文是否支持该主张，它的置信度可以把这条引用标记出来供人工复核。

大语言模型（LLM）在回答问题时附上引用：对每个主张，给出一段源文档的小节以及它所依据的引文。其中一些引用是错误的或凭空捏造的：引文可能根本不在文档中，也可能逐字出现在文档里，但其上下文说的却与主张相反。

人工核对一条引用很慢：找到文档，在文档里找到引文，然后读足够多的上下文来判断它是否支持该主张。

为了自动完成这项核对，我们首先用普通的字符串匹配找出缺失的引文，然后用一个 Choice 问题读取每一条幸存引文的上下文，判断它是否支持该主张。

%%{init: {"flowchart": {"wrappingWidth": 330}}}%% flowchart LR cite["源文档 + 引用"] match{"引文 是否在源文档中？"} fab["标记为 fabricated"] subgraph request[" "] q["Choice — 这一小节与主张是 什么关系？ supports → 标记为 verified contradicts → 标记为 contradicted says nothing → 标记为 unsupported"] end gate{"置信度 ≥ 0.8？"} stand["让判定结果生效"] review["由人工确认"] cite --> match %% the two edges that reach the call come first, so they stay adjacent; the %% string match's own verdict is declared last and lands below them match -- "found" --> request match -- "no quote" --> request match -- "not found" --> fab request --> gate gate --> stand gate --> review

下面，来自一个 LLM 关于 RFC 7519（JSON Web Token）的回答的八条引用经过这项检查。其中四条准确的引用以 0.93 或更高的置信度返回 verified。四个人为植入的失败全部被捕捉到：一条捏造的引文、一个被反驳的主张，以及两条被送去人工处理的不受支持引用。

check_citation() 是你在这里构建的函数，它接收一份源文档和一条引用，返回四种判定之一：verified、unsupported、contradicted 或 fabricated。它还会返回一个置信度，用来标记出需要人工查看的那些。

环境准备

bash

pip install ipython "typesafe-sdk>=0.5.7" cooksafe --extra-index-url https://pypi.typesafe.ai/

然后设置 TYPESAFE_API_KEY。每次 API 调用都会缓存到 json_cache.json，该文件随实践手册一起提供，因此重新运行时会重放已发布的数字，而不是调用 API。删除该文件即可全部实时运行。

下面的数字来自 2026-08-16 的 jev-1.12。

python

import json
import os
import re
from pathlib import Path
from time import perf_counter
 
from cooksafe import JsonCache, make_playground_link
from IPython.display import Markdown, display
from typesafe_sdk import Choice, TypeSafeClient
 
TYPESAFE_MODEL = "jev-1.12"
AUTO_ACCEPT = 0.8  # start high for more human review as you build trust in the model
 
client = TypeSafeClient(
    api_key=os.environ.get("TYPESAFE_API_KEY", "cache-only"),
    base_url=os.environ.get("TYPESAFE_ENDPOINT"),
    timeout=120.0,
)
json_cache = JsonCache(Path("json_cache.json"))

加载源文档与引用

源文档是 RFC 7519（JSON Web Token），从 rfc-editor.org 获取，并作为 rfc7519.txt 提交在本实践手册旁边。下面的代码去掉页眉和页脚，然后把文本切分为带编号的小节。

citations.json 中的八条引用由一个 LLM 针对该 RFC 撰写。四条是准确的；另外四条被我们改成了无法通过检查。

python

def load_source() -> str:
    """RFC 7519 verbatim, minus the page headers and footers that interrupt its paragraphs."""
    lines = []
    for line in Path("rfc7519.txt").read_text().splitlines():
        bare = line.lstrip("\f")
        if re.match(r"Jones, et al\.\s.*\[Page \d+\]$", bare):
            continue
        if re.match(r"RFC 7519\s+JSON Web Token \(JWT\)\s+May 2015$", bare):
            continue
        lines.append(bare)
    return re.sub(r"\n{3,}", "\n\n", "\n".join(lines))
 
 
def split_sections(source: str) -> dict[str, str]:
    """Map each numbered section ("4.1.3") to its text, split on the RFC's header lines."""
    boundary = re.compile(r"(?m)^(?:(\d+(?:\.\d+)*)\.  .+|Appendix [A-Z]\..*)$")
    marks = list(boundary.finditer(source))
    sections = {}
    for mark, nxt in zip(marks, marks[1:] + [None]):
        if mark.group(1) is None:  # an appendix header only terminates the section before it
            continue
        sections[mark.group(1)] = source[mark.start() : nxt.start() if nxt else len(source)].strip()
    return sections
 
 
SOURCE = load_source()
SECTIONS = split_sections(SOURCE)
CITATIONS = json.loads(Path("citations.json").read_text())
 
print(f"{len(SOURCE):,} characters, {len(SECTIONS)} numbered sections, {len(CITATIONS)} citations")
print("\nA citation with a quote:")
print(json.dumps(CITATIONS[1], indent=2))
print("\nA claim-only citation:")
print(json.dumps(next(c for c in CITATIONS if c["quote"] is None), indent=2))

text

58,365 characters, 45 numbered sections, 8 citations
 
A citation with a quote:
{
  "id": "aud_reject",
  "claim": "If a validator does not find itself in a token's audience list, it has to reject the token.",
  "quote": "If the principal processing the claim does not identify itself with a value in the \"aud\" claim when this claim is present, then the JWT MUST be rejected.",
  "section": "4.1.3"
}
 
A claim-only citation:
{
  "id": "iat_future",
  "claim": "The \"iat\" claim requires validators to reject tokens whose issue time is in the future.",
  "quote": null,
  "section": "4.1.6"
}

在源文档中查找每条引文

不在源文档中的引文就是捏造的，发现这一点不需要任何模型。对空白和弯引号做归一化，使引文在 RFC 的换行之间仍能匹配，然后把它作为子串查找。匹配结果还能说明引文来自哪一小节，而那一小节就是下一步模型要读的文本。

一条引用可以只指明某个小节而不从其中引用任何内容。这种情况下没有什么可匹配的，因此直接取该引用所指明的小节，交给模型。

python

def normalize(text: str) -> str:
    """Collapse whitespace and fold curly quotes, so a quote matches across line wraps."""
    table = str.maketrans({"“": '"', "”": '"', "‘": "'", "’": "'"})
    return re.sub(r"\s+", " ", text.translate(table)).strip()
 
 
def find_quote(sections: dict[str, str], quote: str) -> str | None:
    """The number of the section that contains the quote verbatim, or None."""
    needle = normalize(quote)
    for number in sorted(sections, key=lambda n: [int(p) for p in n.split(".")]):
        if needle in normalize(sections[number]):
            return number
    return None
 
 
def locate(sections: dict[str, str], citation: dict) -> tuple[str, str | None]:
    """Step 1 for one citation: a status, plus the section step 2 will read."""
    if citation["quote"] is None:
        return "section-only", sections[citation["section"]]
    number = find_quote(sections, citation["quote"])
    if number is None:
        return "missing", None
    return "found", sections[number]
 
 
for citation in CITATIONS:
    status, section = locate(SECTIONS, citation)
    where = f"section of {len(section):,} chars" if section else "not in the source"
    print(f"{citation['id']:<18}{status:<14}{where}")

text

epoch_seconds     found         section of 3,122 chars
aud_reject        found         section of 761 chars
sig_reporting     missing       not in the source
clock_skew        found         section of 529 chars
exp_required      found         section of 529 chars
pii_encryption    found         section of 1,653 chars
iat_future        section-only  section of 270 chars
duplicate_names   found         section of 918 chars

验证源文档是否支持该主张

到这里仍然带有引文的引用，都与源文档逐字匹配。但这还不够：引文可能准确，而建立在它之上的主张仍然错误。要判断这一点，需要引文的上下文，也就是第 1 步找到的那一小节。

对每条幸存的引用问一个 Choice 问题，涵盖小节与主张之间的三种关系。概率最高的选项就是判定结果，而 AUTO_ACCEPT（上面代码中的 0.8）决定如何处理它：

置信度达到或高于 0.8：判定结果自行生效；
低于 0.8：在有任何东西依据该判定行动之前，由人工确认。

先把阈值设高，随着你观察到模型在自己的文档上的表现，再逐步调低。

python

QUESTIONS = {
    "relation": Choice(
        instructions="How does the section relate to the claim?",
        criteria={
            "supports": "The section states the claim or directly implies that it is true",
            "contradicts": "The section states the opposite of the claim or implies it is false",
            "says_nothing": "The section does not address what the claim asserts, either way",
        },
    ),
}
 
RELATION_TO_VERDICT = {
    "supports": "verified",
    "contradicts": "contradicted",
    "says_nothing": "unsupported",
}
 
 
@json_cache
def ask(claim: str, section: str) -> dict:
    started = perf_counter()
    response = client.system_one(
        state={"claim": claim, "section": section},
        questions=QUESTIONS,
        model=TYPESAFE_MODEL,
    )
    answer = response.answers["relation"]
    return {
        "choice": answer.choice,
        "probabilities": answer.probabilities,
        "confidence": answer.confidence,
        "seconds": round(perf_counter() - started, 2),
        "input_tokens": response.usage.input_tokens or 0,
        "output_tokens": response.usage.output_tokens or 0,
    }
 
 
def verdict(status: str, answer: dict | None) -> dict:
    """Fold step 1 and step 2 into one of the four labels, plus an auto-or-review flag."""
    if status == "missing":
        # confidence None: no model was called, so there is no model confidence to report
        return {"verdict": "fabricated", "confidence": None, "auto": True}
    return {
        "verdict": RELATION_TO_VERDICT[answer["choice"]],
        "confidence": answer["confidence"],
        "auto": answer["confidence"] >= AUTO_ACCEPT,
    }
 
 
def check_citation(sections: dict[str, str], citation: dict) -> dict:
    status, section = locate(sections, citation)
    answer = ask(citation["claim"], section) if section is not None else None
    return {"id": citation["id"], "status": status, "answer": answer, **verdict(status, answer)}

检查每一条引用

八条引用全部经过同一项检查：

python

print(f"{'citation':<18}{'quote':<14}{'relation':<14}{'conf':>6}  {'verdict':<13}{'action':>7}")
for citation in CITATIONS:
    result = check_citation(SECTIONS, citation)
    answer = result["answer"]
    relation = answer["choice"] if answer else "-"
    conf = f"{answer['confidence']:.2f}" if answer else "-"
    action = "auto" if result["auto"] else "review"
    print(
        f"{result['id']:<18}{result['status']:<14}{relation:<14}{conf:>6}"
        f"  {result['verdict']:<13}{action:>7}"
    )

text

citation          quote         relation        conf  verdict       action
epoch_seconds     found         supports        0.93  verified        auto
aud_reject        found         supports        0.95  verified        auto
sig_reporting     missing       -                  -  fabricated      auto
clock_skew        found         supports        0.99  verified        auto
exp_required      found         contradicts     0.99  contradicted    auto
pii_encryption    found         says_nothing    0.27  unsupported   review
iat_future        section-only  says_nothing    0.56  unsupported   review
duplicate_names   found         supports        0.99  verified        auto

四条引用返回 verified，一条返回 fabricated，一条返回 contradicted，两条返回 unsupported。

epoch_seconds、aud_reject、clock_skew 和 duplicate_names 就是那四条准确的引用。它们全部以 0.93 或更高的置信度返回 verified，远高于 AUTO_ACCEPT。
sig_reporting 从未到达模型。它的引文不在 RFC 中，所以仅凭字符串匹配就把它标记为 fabricated。
exp_required 逐字引用了第 4.1.4 节，而同一节写着「Use of this claim is OPTIONAL」，所以它是 contradicted，置信度为 0.99。
pii_encryption 和 iat_future 以 0.27 和 0.56 返回 unsupported，两者都低于阈值，因此都交给了人工。pii_encryption 说明了为什么仅靠字符串匹配还不够：它的引文逐字出现在源文档中，而它来自的那一小节对主张什么也没说。

要把它用于你自己的数据，请替换 rfc7519.txt 和 citations.json。 load_source() 和 split_sections() 是针对 RFC 的版式编写的，因此其他形态的文档需要自己的解析逻辑。

归一化之后字符串匹配是精确的：被截断或略微改写的引文都会返回 fabricated。若要容忍不严谨的引用方式，生产系统需要改用模糊匹配。

在 playground 中打开它

该链接包含一条引用的主张和小节，以及那个问题。打开它即可在浏览器中实时运行同样的调用。

python

example = next(c for c in CITATIONS if c["id"] == "exp_required")
_, example_section = locate(SECTIONS, example)
playground_link = make_playground_link(
    {"claim": example["claim"], "section": example_section}, QUESTIONS, models=[TYPESAFE_MODEL]
)
display(Markdown(f"🔗 [Open one citation's claim + section in the TypeSafe playground]({playground_link})"))

在 TypeSafe playground 中打开一条引用的主张 + 小节 →

复核引用

#环境准备

#加载源文档与引用

#在源文档中查找每条引文

#验证源文档是否支持该主张

#检查每一条引用

#在 playground 中打开它

环境准备

加载源文档与引用

在源文档中查找每条引文

验证源文档是否支持该主张

检查每一条引用

在 playground 中打开它