包裹，长文用
/
划分小节。 3. 列表内容用
- ，引用用
，代码用
，重要句子可用。
4. 保留原文所有信息，不得增删改内容。
5. 不要输出//等文档级标签。
正文是：
"Recurring cash flows are the backbone of cash-flow underwriting: they separate durable revenue from one-off deposits, expose standing obligations, and ultimately determine ability to pay. For small businesses, recurrence is both indispensable and hard to estimate, because much of their income arrives as ambiguous transfers, most payment streams are short, and schedules follow calendar conventions that frequency-domain methods treat as noise. We present CADENCE, Calendar-Aware Distributional Event Network for Cash-Flow Estimation, a compact causal Transformer that treats each transaction as a token and predicts the amount and a full, nonparametric timing distribution of every cash flow's next payment. Inheriting transaction semantics from an external labeler and a frozen text encoder keeps CADENCE compact yet general across the long tail of small-business counterparties, while predicting a calibrated distribution rather than an opaque embedding makes its outputs interpretable enough to feed existing underwriting models and power credit decisions. Ablations confirm the role of each design choice, with counterparty semantics driving recall on short histories, calendar inputs driving timing accuracy, and no-event supervision driving precision. Against the production FFT baseline, CADENCE improves one-day timing accuracy by 93.3% and next-event F1 by 6.7%, while producing calibrated distributions that serve directly as a confidence gate for downstream decisions. In an end-to-end credit evaluation on production applications, replacing FFT recurrence features with CADENCE features raises KS by 4.8%, raises the approval rate at a fixed loss rate by 18.2%, and lowers the loss rate at a fixed approval rate by 13.1%."
我需要分析这个正文。它看起来是一个技术新闻或研究摘要，讨论CADENCE系统。
现在，根据规则：
- 每个自然段落用包裹。我需要识别自然段落。
看正文：
- 第一句："Recurring cash flows are the backbone..." 这似乎是第一段的开始。
- 然后："For small businesses, recurrence is both indispensable..." 这可能继续第一段或新段落。
- 接着："We present CADENCE..." 这可能是新段落。
- 然后："Inheriting transaction semantics..." 可能继续。
- "Ablations confirm..." 新段落。
- "Against the production FFT baseline..." 新段落。
- "In an end-to-end credit evaluation..." 新段落。
我应该将正文分成自然段落。通常，在新闻中，每个句子或一组相关句子形成一个段落。但这里没有明显的段落分隔符。用户提供的正文是连续的文本。
为了符合规则，我需要将整个文本分成逻辑段落。让我试着分组：
1. 介绍 recurring cash flows 的重要性。
2. 讨论小企业的困难。
3. 介绍 CADENCE。
4. 描述 CADENCE 的设计。
5. 讨论消融研究。
6. 与 FFT 基线比较。
7. 端到端信用评估结果。
所以，我可以分成几个段落：
- 第一段：Recurring cash flows are the backbone... determine ability to pay.
- 第二段：For small businesses... treat as noise.
- 第三段：We present CADENCE... next payment.
- 第四段：Inheriting transaction semantics... credit decisions.
- 第五段：Ablations confirm... precision.
- 第六段：Against the production FFT baseline... downstream decisions.
- 第七段：In an end-to-end credit evaluation... 13.1%.
规则说“每个自然段落用
包裹”，所以我需要将每个逻辑段落用
标签包裹。
另外，规则提到“长文用
/
划分小节”。但这个正文不是很长，可能不需要
或
。用户说“长文”，但这里只有一段文本。我应该检查是否有必要划分小节。
看内容：它讨论了一个主题，CADENCE，但结构上没有明显的子部分。可能不需要
或
。规则说“长文用
/
划分小节”，但这里可能不算很长，所以或许只用
。
为了安全，我可以假设整个是一个段落，但分成多个
。
规则4：保留原文所有信息，不得增删改内容。所以我必须保持文本原样。
规则3：列表内容用
- ，引用用
，代码用
，重要句子可用。
在正文中，有没有列表、引用、代码？
- 没有明显的列表。列表通常是项目符号或编号。
- 没有引用（如“as said by...”）。
- 没有代码块。
- 重要句子：可能有些句子可以加，比如关键发现。
规则说“重要句子可用”，意思是我可以选择加到重要句子，但不是必须。我需要判断哪些是重要句子。
例如：
- "CADENCE improves one-day timing accuracy by 93.3% and next-event F1 by 6.7%" – 这可能是重要数据。
- "raises KS by 4.8%, raises the approval rate at a fixed loss rate by 18.2%, and lowers the loss rate at a fixed approval rate by 13.1%" – 重要。
但我不应该添加新内容；只是用标记重要部分。 规则说“可用”，意思是可以选择使用。 现在，关于段落划分： 让我写出正文并分成段落： 原文： "Recurring cash flows are the backbone of cash-flow underwriting: they separate durable revenue from one-off deposits, expose standing obligations, and ultimately determine ability to pay. For small businesses, recurrence is both indispensable and hard to estimate, because much of their income arrives as ambiguous transfers, most payment streams are short, and schedules follow calendar conventions that frequency-domain methods treat as noise. We present CADENCE, Calendar-Aware Distributional Event Network for Cash-Flow Estimation, a compact causal Transformer that treats each transaction as a token and predicts the amount and a full, nonparametric timing distribution of every cash flow's next payment. Inheriting transaction semantics from an external labeler and a frozen text encoder keeps CADENCE compact yet general across the long tail of small-business counterparties, while predicting a calibrated distribution rather than an opaque embedding makes its outputs interpretable enough to feed existing underwriting models and power credit decisions. Ablations confirm the role of each design choice, with counterparty semantics driving recall on short histories, calendar inputs driving timing accuracy, and no-event supervision driving precision. Against the production FFT baseline, CADENCE improves one-day timing accuracy by 93.3% and next-event F1 by 6.7%, while producing calibrated distributions that serve directly as a confidence gate for downstream decisions. In an end-to-end credit evaluation on production applications, replacing FFT recurrence features with CADENCE features raises KS by 4.8%, raises the approval rate at a fixed loss rate by 18.2%, and lowers the loss rate at a fixed approval rate by 13.1%." 分成自然段落： - 第一句和第二句似乎相关：介绍问题。 "Recurring cash flows are the backbone... treat as noise." – 这可以是一个段落。 - 第三和第四句：介绍CADENCE和设计。 "We present CADENCE... credit decisions." – 另一个段落。 - 第五句：消融研究。 "Ablations confirm... precision." – 段落。
，代码用，重要句子可用。 4. 保留原文所有信息，不得增删改内容。 5. 不要输出//等文档级标签。 正文是： "Recurring cash flows are the backbone of cash-flow underwriting: they separate durable revenue from one-off deposits, expose standing obligations, and ultimately determine ability to pay. For small businesses, recurrence is both indispensable and hard to estimate, because much of their income arrives as ambiguous transfers, most payment streams are short, and schedules follow calendar conventions that frequency-domain methods treat as noise. We present CADENCE, Calendar-Aware Distributional Event Network for Cash-Flow Estimation, a compact causal Transformer that treats each transaction as a token and predicts the amount and a full, nonparametric timing distribution of every cash flow's next payment. Inheriting transaction semantics from an external labeler and a frozen text encoder keeps CADENCE compact yet general across the long tail of small-business counterparties, while predicting a calibrated distribution rather than an opaque embedding makes its outputs interpretable enough to feed existing underwriting models and power credit decisions. Ablations confirm the role of each design choice, with counterparty semantics driving recall on short histories, calendar inputs driving timing accuracy, and no-event supervision driving precision. Against the production FFT baseline, CADENCE improves one-day timing accuracy by 93.3% and next-event F1 by 6.7%, while producing calibrated distributions that serve directly as a confidence gate for downstream decisions. In an end-to-end credit evaluation on production applications, replacing FFT recurrence features with CADENCE features raises KS by 4.8%, raises the approval rate at a fixed loss rate by 18.2%, and lowers the loss rate at a fixed approval rate by 13.1%." 我需要分析这个正文。它看起来是一个技术新闻或研究摘要，讨论CADENCE系统。 现在，根据规则： - 每个自然段落用包裹。我需要识别自然段落。 看正文： - 第一句："Recurring cash flows are the backbone..." 这似乎是第一段的开始。 - 然后："For small businesses, recurrence is both indispensable..." 这可能继续第一段或新段落。 - 接着："We present CADENCE..." 这可能是新段落。 - 然后："Inheriting transaction semantics..." 可能继续。 - "Ablations confirm..." 新段落。 - "Against the production FFT baseline..." 新段落。 - "In an end-to-end credit evaluation..." 新段落。 我应该将正文分成自然段落。通常，在新闻中，每个句子或一组相关句子形成一个段落。但这里没有明显的段落分隔符。用户提供的正文是连续的文本。 为了符合规则，我需要将整个文本分成逻辑段落。让我试着分组： 1. 介绍 recurring cash flows 的重要性。 2. 讨论小企业的困难。 3. 介绍 CADENCE。 4. 描述 CADENCE 的设计。 5. 讨论消融研究。 6. 与 FFT 基线比较。 7. 端到端信用评估结果。 所以，我可以分成几个段落： - 第一段：Recurring cash flows are the backbone... determine ability to pay. - 第二段：For small businesses... treat as noise. - 第三段：We present CADENCE... next payment. - 第四段：Inheriting transaction semantics... credit decisions. - 第五段：Ablations confirm... precision. - 第六段：Against the production FFT baseline... downstream decisions. - 第七段：In an end-to-end credit evaluation... 13.1%. 规则说“每个自然段落用
包裹”，所以我需要将每个逻辑段落用
标签包裹。 另外，规则提到“长文用
/
划分小节”。但这个正文不是很长，可能不需要
或
。用户说“长文”，但这里只有一段文本。我应该检查是否有必要划分小节。 看内容：它讨论了一个主题，CADENCE，但结构上没有明显的子部分。可能不需要
或
。规则说“长文用
/
划分小节”，但这里可能不算很长，所以或许只用
。 为了安全，我可以假设整个是一个段落，但分成多个
。 规则4：保留原文所有信息，不得增删改内容。所以我必须保持文本原样。 规则3：列表内容用
- ，引用用
，代码用，重要句子可用。 在正文中，有没有列表、引用、代码？ - 没有明显的列表。列表通常是项目符号或编号。 - 没有引用（如“as said by...”）。 - 没有代码块。 - 重要句子：可能有些句子可以加，比如关键发现。 规则说“重要句子可用”，意思是我可以选择加到重要句子，但不是必须。我需要判断哪些是重要句子。 例如： - "CADENCE improves one-day timing accuracy by 93.3% and next-event F1 by 6.7%" – 这可能是重要数据。 - "raises KS by 4.8%, raises the approval rate at a fixed loss rate by 18.2%, and lowers the loss rate at a fixed approval rate by 13.1%" – 重要。 但我不应该添加新内容；只是用标记重要部分。 规则说“可用”，意思是可以选择使用。 现在，关于段落划分： 让我写出正文并分成段落： 原文： "Recurring cash flows are the backbone of cash-flow underwriting: they separate durable revenue from one-off deposits, expose standing obligations, and ultimately determine ability to pay. For small businesses, recurrence is both indispensable and hard to estimate, because much of their income arrives as ambiguous transfers, most payment streams are short, and schedules follow calendar conventions that frequency-domain methods treat as noise. We present CADENCE, Calendar-Aware Distributional Event Network for Cash-Flow Estimation, a compact causal Transformer that treats each transaction as a token and predicts the amount and a full, nonparametric timing distribution of every cash flow's next payment. Inheriting transaction semantics from an external labeler and a frozen text encoder keeps CADENCE compact yet general across the long tail of small-business counterparties, while predicting a calibrated distribution rather than an opaque embedding makes its outputs interpretable enough to feed existing underwriting models and power credit decisions. Ablations confirm the role of each design choice, with counterparty semantics driving recall on short histories, calendar inputs driving timing accuracy, and no-event supervision driving precision. Against the production FFT baseline, CADENCE improves one-day timing accuracy by 93.3% and next-event F1 by 6.7%, while producing calibrated distributions that serve directly as a confidence gate for downstream decisions. In an end-to-end credit evaluation on production applications, replacing FFT recurrence features with CADENCE features raises KS by 4.8%, raises the approval rate at a fixed loss rate by 18.2%, and lowers the loss rate at a fixed approval rate by 13.1%." 分成自然段落： - 第一句和第二句似乎相关：介绍问题。 "Recurring cash flows are the backbone... treat as noise." – 这可以是一个段落。 - 第三和第四句：介绍CADENCE和设计。 "We present CADENCE... credit decisions." – 另一个段落。 - 第五句：消融研究。 "Ablations confirm... precision." – 段落。