Skills Plugins MCP Prompt Model 博客 我的中心
Content Creation #data #writing

write-calibration-section

写论文 Calibration / Data section:参数表(带来源) + 稳态矩 + 数据矩对照。遵守 academic-writing-style。当用户说"写校准这一节"、"draft Section 3"、"写参数表"时调用。

DeepseekModel Curated skill Quality Good · 48 v1.0.0

Get

https://deepseekmodel.com/api/download.php?id=allblueuk-structural-modeling-for-economists-skills-write-calibration-section-skill-md&format=skill
Download .skill Standard format with system_prompt and model_config, ready for any agent framework
The actual content of the system_prompt field in the .skill file.
name write-calibration-section description 写论文 Calibration / Data section:参数表(带来源) + 稳态矩 + 数据矩对照。遵守 academic-writing-style。当用户说"写校准这一节"、"draft Section 3"、"写参数表"时调用。 type paper-composition write-calibration-section — 校准章节 触发场景 "写 Calibration section"、"draft Section 3" 校准表完成后 输入契约 字段 说明 paper_dir 论文目录 calibration_csv code/<model>/calibration.csv (每参数有来源) ss_checkpoint results/checkpoints/<model>_ss.json target_moments 来自 spec.md §9.1 的目标矩 前置条件 校准表通过 calibrate-from-moments + math-reviewer 稳态通过 validate-steady-state 步骤清单 Step 1. 子小节结构 3.1 Parameter Calibration :分组(preferences / technology / shocks),每组一个 LaTeX 表( booktabs ),列出参数符号、取值、含义、来源。 3.2 Steady-State Properties :K/Y、C/Y、I/Y 等关键稳态比率,与数据对照。 3.3 Data (若适用):若涉及微观或时间序列数据,描述来源、覆盖期、清洗步骤。否则跳过。 Step 2. 构造 LaTeX 参数表 直接从 calibration.csv 构造: \begin{table}[t] \centering \caption{Calibrated parameters (quarterly).} \label{tab:calibration} \begin{tabular}{lcll} \toprule Parameter & Value & Description & Source \\ \midrule $\beta$ & 0.99 & discount factor & \citet{KingRebelo1999} \\ $\alpha$ & 0.33 & capital share & NIPA, 1990--2019 avg. \\ $\delta$ & 0.025 & quarterly depreciation & 10\% annual rate \\ $\rho_z$ & 0.95 & TFP persistence & \citet{KingRebelo1999} baseline \\ $\sigma_z$ & 0.007 & TFP innovation std.\ & post-war US Solow residual \\ \bottomrule \end{tabular} \end{table} 引用 key 必须存在于 references.bib ,否则调 manage-bibliography 补加。 Step 3. 写稳态对照 \subsection{Steady-State Properties} At the calibrated parameters the model implies $K/Y = 9.40$ at quarterly frequency (equivalently $2.35$ at annual frequency), $C/Y = 0.77$, and $I/Y = 0.23$. These ratios match standard quarterly US business-cycle moments (see, e.g., \citealp{KingRebelo1999}, Table~1). 数字必须来自 checkpoint,不是手写。 Step 4. 写 sections/03_calibration.tex \section{Calibration} \label{sec:calibration} \subsection{Parameter Values and Sources} ... \subsection{Steady-State Properties} ... \subsection{Data} % 若适用 ... Step 5. 评审 paper-reviewer 检查:每个参数表行都有 source 列;引用 key 全部解析;数字与 checkpoint 一致。 输出契约 paper/sections/03_calibration.tex paper/tables/<model>_calibration.tex (可被 main.tex \input{} 的独立表) 评审报告 不允许 列参数但 Source 列写 "calibrated to literature"(含混)— 必须具体到 \citet{} 或 "X data, period Y" 数字与 checkpoint 不一致(手抄错)— 必须脚本化提取 写 "see calibration.csv"(暴露文件路径) 失败回退 症状 处理 参数无 source 阻断,先调 calibrate-from-moments 补来源 数据期与文献频率不符 在 §3.3 显式说明,并由 model-reviewer 重审
Keywords that activate this skill. Click one to copy it.

This skill does not provide trigger words.

The downloaded .skill package contains the following fields.
Field Description
formatFormat tag (skill/v1)
skill_idUnique skill ID
nameSkill name
versionVersion
descriptionDescription
categoryCategories (array)
trigger_wordsTrigger words
tagsTags
sourceSource
source_urlSource URL (this page)
exported_atExported at (set per download)
system_promptSystem prompt body
model_configModel config: provider / model / temperature / max_tokens / top_p
examplesExamples
install_guideImport guide for Coze / Dify / Claude / custom frameworks
The same skill can be exported in different platform formats.
.skill Standard format with system_prompt and model_config, ready for any agent framework Download
.skillpro Enhanced format with scripts, tools, dependencies and hooks Download
.json Plain JSON export with system_prompt and model parameters only Download
Coze Markdown with frontmatter, for Coze platform import Download
Dify Dify DSL, import directly after creating an app Download

每日精选 Skill 推荐,免费送到你邮箱

输入邮箱,每天接收一个精选 AI Agent 技能推荐。完全免费,持续更新。

提交后我们会发送一封确认邮件,点击邮件里的链接才会开始收信。

完全免费,取消任意时间。我们不会发送垃圾邮件。