// 学术论文写作助手，专门用于 LaTeX 论文编写、BibTeX 管理、格式化、学术写作规范检查。适用于 AI/ML 研究论文、会议投稿（NeurIPS、ICML、ICLR 等）
| name | academic-paper-helper |
| description | 学术论文写作助手，专门用于 LaTeX 论文编写、BibTeX 管理、格式化、学术写作规范检查。适用于 AI/ML 研究论文、会议投稿（NeurIPS、ICML、ICLR 等） |
| disable-model-invocation | true |
专为 AI/ML 博士生设计的学术写作助手，提供端到端的论文写作支持。
使用场景：创建新论文时
模板类型：
生成内容：
\documentclass{article}
% 会议特定包
\usepackage[final]{neurips_2024} % 或其他会议
% 标准学术包
\usepackage{amsmath,amssymb,amsfonts}
\usepackage{algorithm,algorithmic}
\usepackage{graphicx}
\usepackage{booktabs}
\usepackage{hyperref}
\title{Your Paper Title}
\author{
Your Name \\
Department \\
University \\
\texttt{email@university.edu}
}
\begin{document}
\maketitle
\begin{abstract}
Your abstract here.
\end{abstract}
\section{Introduction}
% Your content
\section{Related Work}
% Your content
\section{Method}
% Your content
\section{Experiments}
% Your content
\section{Conclusion}
% Your content
\bibliographystyle{plain}
\bibliography{references}
\end{document}
使用场景：管理论文引用
功能：
示例：
输入混乱的论文信息：
attention is all you need
vaswani et al
2017
neurips
输出规范的 BibTeX：
@inproceedings{vaswani2017attention,
title={Attention Is All You Need},
author={Vaswani, Ashish and Shazeer, Noam and Parmar, Niki and Uszkoreit, Jakob and Jones, Llion and Gomez, Aidan N. and Kaiser, {\L}ukasz and Polosukhin, Illia},
booktitle={Advances in Neural Information Processing Systems},
volume={30},
year={2017},
doi={10.5555/3295222.3295349}
}
检查项目：
常见错误及解决：
% 错误 1：Undefined control sequence
% 原因：包未导入或命令拼写错误
% 解决：检查 \usepackage{} 和命令名
% 错误 2：Missing $ inserted
% 原因：数学符号在文本模式中使用
% 解决：使用 $...$ 或 \( ... \) 包围数学内容
% 错误 3：File not found
% 原因：图片或 .bib 文件路径错误
% 解决：检查文件路径，使用相对路径
% 错误 4：Citation undefined
% 原因：.bib 文件未编译或引用键错误
% 解决：运行 bibtex，检查引用键
\section{Introduction}
% 段落 1：研究背景和动机
Recent advances in [field] have led to...
However, existing approaches suffer from...
% 段落 2：具体问题
Consider the problem of [specific problem]...
Current methods [limitation 1], [limitation 2]...
% 段落 3：本文方法概述
To address these challenges, we propose [method name]...
Our approach [key innovation 1], [key innovation 2]...
% 段落 4：主要贡献
The main contributions of this work are:
\begin{itemize}
\item We propose [contribution 1]...
\item We demonstrate [contribution 2]...
\item We achieve [contribution 3]...
\end{itemize}
\section{Method}
\subsection{Problem Formulation}
Let $\mathcal{X}$ denote... We aim to learn...
\subsection{Architecture}
Our model consists of... Figure~\ref{fig:architecture} illustrates...
\subsection{Training Procedure}
We optimize the following objective:
\begin{equation}
\mathcal{L} = \mathbb{E}_{(x,y)\sim\mathcal{D}}[\ell(f_\theta(x), y)]
\end{equation}
\subsection{Implementation Details}
We implement our method in PyTorch...
\section{Experiments}
\subsection{Experimental Setup}
\textbf{Datasets:} We evaluate on...
\textbf{Baselines:} We compare against...
\textbf{Metrics:} We report...
\textbf{Implementation:} We use...
\subsection{Main Results}
Table~\ref{tab:main_results} shows...
\subsection{Ablation Study}
To understand the contribution of each component...
\subsection{Analysis}
Figure~\ref{fig:analysis} visualizes...
\begin{algorithm}
\caption{Your Algorithm}
\label{alg:youralgorithm}
\begin{algorithmic}[1]
\REQUIRE Input $x$
\ENSURE Output $y$
\STATE Initialize $\theta$
\FOR{$t = 1$ to $T$}
\STATE Compute $\nabla_\theta \mathcal{L}$
\STATE Update $\theta \leftarrow \theta - \alpha \nabla_\theta \mathcal{L}$
\ENDFOR
\RETURN $\theta$
\end{algorithmic}
\end{algorithm}
\begin{table}[t]
\centering
\caption{Comparison with baselines}
\label{tab:results}
\begin{tabular}{lcc}
\toprule
Method & Accuracy & F1-Score \\
\midrule
Baseline 1 & 85.3 & 82.1 \\
Baseline 2 & 87.6 & 84.5 \\
\textbf{Ours} & \textbf{91.2} & \textbf{89.3} \\
\bottomrule
\end{tabular}
\end{table}
\begin{figure}[t]
\centering
\includegraphics[width=0.8\linewidth]{figures/architecture.pdf}
\caption{Overall architecture of our model.}
\label{fig:architecture}
\end{figure}
% 行内公式
The loss function $\mathcal{L}(\theta)$ is defined as...
% 单行公式
\begin{equation}
\mathcal{L}(\theta) = \sum_{i=1}^{n} \ell(f_\theta(x_i), y_i)
\end{equation}
% 多行公式
\begin{align}
\mathcal{L}(\theta) &= \mathbb{E}_{x\sim p_{\text{data}}}[\log p_\theta(x)] \\
&= \sum_{i=1}^{n} \log p_\theta(x_i)
\end{align}
正确引用方式：
% 作为名词
\citet{vaswani2017attention} proposed the Transformer architecture.
% 输出：Vaswani et al. (2017) proposed...
% 括号中
The Transformer architecture~\citep{vaswani2017attention} has...
% 输出：... architecture (Vaswani et al., 2017) has...
% 多个引用
Recent work~\citep{vaswani2017attention,devlin2018bert} has shown...
% 输出：... work (Vaswani et al., 2017; Devlin et al., 2018) has...
版本: v1.0 创建日期: 2025-10-21 适用对象: AI/ML 博士生、研究人员 核心价值: 提高学术写作效率，确保论文质量和规范性
Use when the user wants multi-agent division of labor for research-led work and the lead should stay on the critical path while 1-2 bounded sidecars handle low-coupling tasks. Do not use this for tiny tasks, fully sequential debugging, or overlapping refactors.
Use when the user asks to review a skill, analyze skill quality, update a skill version, or run a repeatable keep/disable/archive decision loop from real failures instead of abstract best practices.
Use when the user explicitly wants to upload a final or near-final PDF to paperreview.ai for an external second opinion. Skip this for local paper critique, which should go through `paper-review-pipeline` first.
Use when a task is ambiguous and needs deep multi-role planning with designer, inspiration, and reviewer perspectives before execution. Skip this for ordinary task plans.
落实 CLAUDE.md / AGENTS.md 中的“主动探索”原则，在向用户提问前自动使用 Grep、Read、Bash、WebSearch 等工具获取信息
Use when you want Claude Code CLI as a second opinion for coding tasks such as design tradeoffs, debugging, or diff review, while keeping Codex as the primary implementer.