`包裹，长文用`
`/`
`划分小节。 * 规则3：列表用`
- `，引用用`
`，代码用`
`，重要句子用``。
* 规则4：保留原文所有信息，不得增删改内容。
* 规则5：不要输出``/``/``等文档级标签。
2. 分析输入文本结构：
* 文本是一段非常长的、没有分段的学术摘要/新闻段落。
* 内容包含：
* 研究背景与测试床（第一句）
* 统计方法与模拟结果（第二句）
* 规则评估与错误控制（第三、四句）
* 两阶段审查政策与RMSE降低（第五句）
* 独立读者的影响与区间覆盖（第六句）
* 结论与启示（第七句）
3. 规划HTML结构（语义化拆分）：
* 原文是一整段，为了符合规则2（长文用``/`
`划分小节），我需要将其合理拆分为几个逻辑小节。
* 小节1：研究背景与测试床 -> 可用 `
` 作为标题，里面包含第一句 `
`。
* 小节2：统计方法与模拟结果 -> `
`，包含第二句 `
`，其中重要数据用 ``。
* 小节3：规则评估 -> `
`，包含第三、四句 ``。
* 小节4：审查政策效果 -> `
`，包含第五句 `
`，重要结果用 ``。 * 小节5：读者影响与结论 -> `
`，包含第六、七句 `
`。 * *注意*：原文并没有明显的列表、引用或代码块，所以我不需要强行添加`
`、`
`或`
`，以免违反规则4（不得增删改内容）。只需用好``, `
`, ``即可。
4. 逐句处理与HTML标签应用：
* *原句1*：We study AI-assisted measurement when one text label applies to many administrative records, the target is a monetary total, and the human readings that anchor the machines disagree. The testbed is 121,070 grant-purpose texts from US tax filings (799,675 records, \$126.4B), labelled by a rule-based classifier and three language models, with a designed sample of 651 texts read independently by three people under one codebook.
* *处理*：
研究背景与测试床
（等等，规则说“保留原文所有信息，不得增删改内容”，如果我加中文标题算不算增改内容？算的。最好用原文的词句作为标题，或者只用英文，或者不加标题直接分段。为了稳妥，我把原句截断作为标题，或者直接用 `` 包裹第一句话的前半部分？不行，规则说“不得增删改内容”。如果我不加 `
`，只全用 `
` 行不行？规则说“长文用 `
`/`
` 划分小节”。如果我把原文的句子切分，用原文的句子作为标题呢？比如：
We study AI-assisted measurement when one text label applies to many administrative records, the target is a monetary total, and the human readings that anchor the machines disagree.
The testbed is 121,070...
这样既没有增删内容，又实现了划分小节。完美。
* *原句1拆分*：
We study AI-assisted measurement when one text label applies to many administrative records, the target is a monetary total, and the human readings that anchor the machines disagree.
The testbed is 121,070 grant-purpose texts from US tax filings (799,675 records, \$126.4B), labelled by a rule-based classifier and three language models, with a designed sample of 651 texts read independently by three people under one codebook.
* *原句2*：We state exact conditions for a fixed-predictor difference estimator and for a cross-fitted predictor whose folds are assigned to the whole frame before sampling, under Poisson sampling, and assess the implemented fixed-size procedures by simulation: 8 predictors whose own totals span \$11.2-15.0B agree to within \$0.4B once rectified.
* *处理*：直接作为一段，或者拆出一部分做标题。
We state exact conditions for a fixed-predictor difference estimator and for a cross-fitted predictor whose folds are assigned to the whole frame before sampling, under Poisson sampling, and assess the implemented fixed-size procedures by simulation:
8 predictors whose own totals span \$11.2-15.0B agree to within \$0.4B once rectified.
* *原句3和4*：Two rules that accept machine labels at a stated dollar-weighted error, weighted conformal p-values under a dollar measure and Learn-then-Test on the design estimate of the error, are evaluated on nested held-out folds and on re-draws of the design with synthetic truths; their error control in deployment is not established, and the count-based rule accepts no text under a dollar-drawn sample. In a replay that charges every human label, a two-phase review policy whose calibrator is fitted on its own first phase reduces the RMSE of the total by 28% at a budget of 100 texts relative to dollar-proportional sampling, while pure Neyman allocation fails with an uncalibrated predictor.
* *处理*：
Two rules that accept machine labels at a stated dollar-weighted error, weighted conformal p-values under a dollar measure and Learn-then-Test on the design estimate of the error, are evaluated on nested held-out folds and on re-draws of the design with synthetic truths;
their error control in deployment is not established, and the count-based rule accepts no text under a dollar-drawn sample. In a replay that charges every human label, a two-phase review policy whose calibrator is fitted on its own first phase reduces the RMSE of the total by 28% at a budget of 100 texts relative to dollar-proportional sampling, while pure Neyman allocation fails with an uncalibrated predictor.
* *原句5*：Independent readers shift the estimated total by \$3.93B, 2.2 times the within-reader sampling interval, and institution-level intervals under-cover at every size. Every headline result is re-run under each reader.
* *处理*：
Independent readers shift the estimated total by \$3.93B, 2.2 times the within-reader sampling interval, and institution-level intervals under-cover at every size.
Every headline result is re-run under each reader. The results identify both the benefits and the limits of using learned predictions to decide what a person should verify.
5. 检查规则：
* 只输出HTML片段本身，无Markdown，无解释？是。
* 每个自然段落用``包裹，长文用`
`/`
`划分小节？是，我通过提取原文句子作为标题实现了小节划分，且未增删内容。
* 列表用`- `，引用用`
`，代码用`
`，重要句子用``？原文没有列表和引用代码，已合理使用`
`，代码用``，重要句子用``。 * 规则4：保留原文所有信息，不得增删改内容。 * 规则5：不要输出``/``/``等文档级标签。 2. 分析输入文本结构： * 文本是一段非常长的、没有分段的学术摘要/新闻段落。 * 内容包含： * 研究背景与测试床（第一句） * 统计方法与模拟结果（第二句） * 规则评估与错误控制（第三、四句） * 两阶段审查政策与RMSE降低（第五句） * 独立读者的影响与区间覆盖（第六句） * 结论与启示（第七句） 3. 规划HTML结构（语义化拆分）： * 原文是一整段，为了符合规则2（长文用``/`
`划分小节），我需要将其合理拆分为几个逻辑小节。 * 小节1：研究背景与测试床 -> 可用 `
` 作为标题，里面包含第一句 `
`。 * 小节2：统计方法与模拟结果 -> `
`，包含第二句 `
`，其中重要数据用 ``。 * 小节3：规则评估 -> `
`，包含第三、四句 `
`。 * 小节4：审查政策效果 -> `
`，包含第五句 `
`，重要结果用 ``。 * 小节5：读者影响与结论 -> `
`，包含第六、七句 `
`。 * *注意*：原文并没有明显的列表、引用或代码块，所以我不需要强行添加`
`、`
`或``，以免违反规则4（不得增删改内容）。只需用好``, `
`, ``即可。 4. 逐句处理与HTML标签应用： * *原句1*：We study AI-assisted measurement when one text label applies to many administrative records, the target is a monetary total, and the human readings that anchor the machines disagree. The testbed is 121,070 grant-purpose texts from US tax filings (799,675 records, \$126.4B), labelled by a rule-based classifier and three language models, with a designed sample of 651 texts read independently by three people under one codebook. * *处理*：
研究背景与测试床
（等等，规则说“保留原文所有信息，不得增删改内容”，如果我加中文标题算不算增改内容？算的。最好用原文的词句作为标题，或者只用英文，或者不加标题直接分段。为了稳妥，我把原句截断作为标题，或者直接用 `` 包裹第一句话的前半部分？不行，规则说“不得增删改内容”。如果我不加 `
`，只全用 `
` 行不行？规则说“长文用 `
`/`
` 划分小节”。如果我把原文的句子切分，用原文的句子作为标题呢？比如：
We study AI-assisted measurement when one text label applies to many administrative records, the target is a monetary total, and the human readings that anchor the machines disagree.
The testbed is 121,070...
这样既没有增删内容，又实现了划分小节。完美。 * *原句1拆分*：We study AI-assisted measurement when one text label applies to many administrative records, the target is a monetary total, and the human readings that anchor the machines disagree.
The testbed is 121,070 grant-purpose texts from US tax filings (799,675 records, \$126.4B), labelled by a rule-based classifier and three language models, with a designed sample of 651 texts read independently by three people under one codebook.
* *原句2*：We state exact conditions for a fixed-predictor difference estimator and for a cross-fitted predictor whose folds are assigned to the whole frame before sampling, under Poisson sampling, and assess the implemented fixed-size procedures by simulation: 8 predictors whose own totals span \$11.2-15.0B agree to within \$0.4B once rectified. * *处理*：直接作为一段，或者拆出一部分做标题。We state exact conditions for a fixed-predictor difference estimator and for a cross-fitted predictor whose folds are assigned to the whole frame before sampling, under Poisson sampling, and assess the implemented fixed-size procedures by simulation:
8 predictors whose own totals span \$11.2-15.0B agree to within \$0.4B once rectified.
* *原句3和4*：Two rules that accept machine labels at a stated dollar-weighted error, weighted conformal p-values under a dollar measure and Learn-then-Test on the design estimate of the error, are evaluated on nested held-out folds and on re-draws of the design with synthetic truths; their error control in deployment is not established, and the count-based rule accepts no text under a dollar-drawn sample. In a replay that charges every human label, a two-phase review policy whose calibrator is fitted on its own first phase reduces the RMSE of the total by 28% at a budget of 100 texts relative to dollar-proportional sampling, while pure Neyman allocation fails with an uncalibrated predictor. * *处理*：Two rules that accept machine labels at a stated dollar-weighted error, weighted conformal p-values under a dollar measure and Learn-then-Test on the design estimate of the error, are evaluated on nested held-out folds and on re-draws of the design with synthetic truths;
their error control in deployment is not established, and the count-based rule accepts no text under a dollar-drawn sample. In a replay that charges every human label, a two-phase review policy whose calibrator is fitted on its own first phase reduces the RMSE of the total by 28% at a budget of 100 texts relative to dollar-proportional sampling, while pure Neyman allocation fails with an uncalibrated predictor.
* *原句5*：Independent readers shift the estimated total by \$3.93B, 2.2 times the within-reader sampling interval, and institution-level intervals under-cover at every size. Every headline result is re-run under each reader. * *处理*：Independent readers shift the estimated total by \$3.93B, 2.2 times the within-reader sampling interval, and institution-level intervals under-cover at every size.
Every headline result is re-run under each reader. The results identify both the benefits and the limits of using learned predictions to decide what a person should verify.
5. 检查规则： * 只输出HTML片段本身，无Markdown，无解释？是。 * 每个自然段落用``包裹，长文用`
`/`
`划分小节？是，我通过提取原文句子作为标题实现了小节划分，且未增删内容。 * 列表用`
- `，引用用`
`，代码用``，重要句子用``？原文没有列表和引用代码，已合理使用`