ml-paper-writing

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

ML Paper Writing for Top AI Conferences

面向顶级 AI 会议的 ML 论文写作

Expert-level guidance for writing publication-ready papers targeting NeurIPS, ICML, ICLR, ACL, AAAI, COLM. This skill combines writing philosophy from top researchers (Nanda, Farquhar, Karpathy, Lipton, Steinhardt) with practical tools: LaTeX templates, citation verification APIs, and conference checklists.

For systems venues (OSDI, NSDI, ASPLOS, SOSP), use the systems-paper-writing skill, which provides paragraph-level structural blueprints, writing patterns, venue-specific checklists, and LaTeX templates for systems conferences.

面向 NeurIPS、ICML、ICLR、ACL、AAAI、COLM 撰写达到发表标准论文的专家级指导。本 skill 将顶级研究者（Nanda、Farquhar、Karpathy、Lipton、Steinhardt）的写作理念与实用工具结合起来，包括 LaTeX 模板、引用验证 API 和会议检查清单。

对于系统类会议（OSDI、NSDI、ASPLOS、SOSP），请使用 systems-paper-writing skill。它为系统类会议提供段落级结构蓝图、写作模式、特定会议检查清单和 LaTeX 模板。

Core Philosophy: Collaborative Writing

核心理念：协作式写作

Paper writing is collaborative, but Claude should be proactive in delivering drafts.

The typical workflow starts with a research repository containing code, results, and experimental artifacts. Claude's role is to:

Understand the project by exploring the repo, results, and existing documentation
Deliver a complete first draft when confident about the contribution
Search literature using web search and APIs to find relevant citations
Refine through feedback cycles when the scientist provides input
Ask for clarification only when genuinely uncertain about key decisions

Key Principle: Be proactive. If the repo and results are clear, deliver a full draft. Don't block waiting for feedback on every section—scientists are busy. Produce something concrete they can react to, then iterate based on their response.

论文写作是协作过程，但 Claude 应主动交付草稿。

典型工作流从一个包含代码、结果和实验产物的研究代码仓库开始。Claude 的职责是：

理解项目：探索代码仓库、结果和现有文档
在对贡献有把握时交付完整初稿
检索文献：使用网络搜索和 API 查找相关引用
通过反馈循环进行完善：当科学家提供意见时持续迭代
仅在确实无法确定关键决策时询问澄清问题

关键原则：保持主动。如果代码仓库和结果都很清晰，就交付完整草稿。不要在每一节都等待反馈，从而阻塞进度——科学家都很忙。先产出可供反馈的具体内容，再根据其反馈进行迭代。

⚠️ CRITICAL: Never Hallucinate Citations

⚠️ 关键要求：绝不能幻觉式生成引用

This is the most important rule in academic writing with AI assistance.

这是 AI 辅助学术写作中最重要的规则。

The Problem

问题

AI-generated citations have a ~40% error rate. Hallucinated references—papers that don't exist, wrong authors, incorrect years, fabricated DOIs—are a serious form of academic misconduct that can result in desk rejection or retraction.

AI 生成的引用约有 40% 的错误率。幻觉引用——包括不存在的论文、错误的作者、不正确的年份和伪造的 DOI——是一种严重的学术不端行为，可能导致编辑直接拒稿或论文撤稿。

The Rule

规则

NEVER generate BibTeX entries from memory. ALWAYS fetch programmatically.

Action	✅ Correct	❌ Wrong
Adding a citation	Search API → verify → fetch BibTeX	Write BibTeX from memory
Uncertain about a paper	Mark as `[CITATION NEEDED]`	Guess the reference
Can't find exact paper	Note: "placeholder - verify"	Invent similar-sounding paper

绝不要凭记忆生成 BibTeX 条目。始终以程序化方式获取。

操作	✅ 正确做法	❌ 错误做法
添加引用	搜索 API → 验证 → 获取 BibTeX	凭记忆编写 BibTeX
对某篇论文不确定	标记为 `[CITATION NEEDED]`	猜测参考文献
找不到完全匹配的论文	注明：“placeholder - verify”	杜撰一篇名称相近的论文

When You Can't Verify a Citation

无法验证引用时

If you cannot programmatically verify a citation, you MUST:

latex

% EXPLICIT PLACEHOLDER - requires human verification
\cite{PLACEHOLDER_author2024_verify_this}  % TODO: Verify this citation exists

Always tell the scientist: "I've marked [X] citations as placeholders that need verification. I could not confirm these papers exist."

如果无法以程序化方式验证引用，则必须：

latex

% EXPLICIT PLACEHOLDER - requires human verification
\\cite{PLACEHOLDER_author2024_verify_this}  % TODO: Verify this citation exists

始终告知科学家：“我已将 [X] 条引用标记为需要验证的占位符。无法确认这些论文确实存在。”

Recommended: Install Exa MCP for Paper Search

推荐：安装 Exa MCP 以搜索论文

For the best paper search experience, install Exa MCP which provides real-time academic search:

Claude Code:

bash

claude mcp add exa -- npx -y mcp-remote "https://mcp.exa.ai/mcp"

Cursor / VS Code (add to MCP settings):

json

{
  "mcpServers": {
    "exa": {
      "type": "http",
      "url": "https://mcp.exa.ai/mcp"
    }
  }
}

Exa MCP enables searches like:

"Find papers on RLHF for language models published after 2023"
"Search for transformer architecture papers by Vaswani"
"Get recent work on sparse autoencoders for interpretability"

Then verify results with Semantic Scholar API and fetch BibTeX via DOI.

如需获得最佳论文搜索体验，请安装 Exa MCP，它提供实时学术搜索：

Claude Code：

bash

claude mcp add exa -- npx -y mcp-remote "https://mcp.exa.ai/mcp"

Cursor / VS Code（添加到 MCP 设置）：

json

{
  "mcpServers": {
    "exa": {
      "type": "http",
      "url": "https://mcp.exa.ai/mcp"
    }
  }
}

Exa MCP 支持以下类型的搜索：

"Find papers on RLHF for language models published after 2023"
"Search for transformer architecture papers by Vaswani"
"Get recent work on sparse autoencoders for interpretability"

然后使用 Semantic Scholar API 验证结果，并通过 DOI 获取 BibTeX。

Workflow 0: Starting from a Research Repository

工作流 0：从研究代码仓库开始

When the user points you at a code repository (results, configs, logs) instead of a draft, follow the repository-exploration workflow in references/from-research-repo.md to extract the contribution, results and existing citations before writing.

当用户指向一个代码仓库（包含结果、配置和日志），而不是提供现成草稿时，应遵循 references/from-research-repo.md 中的代码仓库探索工作流，在写作前提取研究贡献、结果和现有引用。

When to Use This Skill

何时使用本 Skill

Use this skill when:

Starting from a research repo to write a paper
Drafting or revising specific sections
Finding and verifying citations for related work
Formatting for conference submission
Resubmitting to a different venue (format conversion)
Iterating on drafts with scientist feedback

Always remember: First drafts are starting points for discussion, not final outputs.

在以下场景使用本 skill：

从研究代码仓库开始撰写论文
起草或修改特定章节
查找并验证相关工作引用
按会议要求进行排版
改投其他会议（转换格式）
根据科学家反馈迭代完善草稿

始终记住：初稿是讨论的起点，不是最终成果。

Balancing Proactivity and Collaboration

在主动性与协作之间取得平衡

Default: Be proactive. Deliver drafts, then iterate.

Confidence Level	Action
High (clear repo, obvious contribution)	Write full draft, deliver, iterate on feedback
Medium (some ambiguity)	Write draft with flagged uncertainties, continue
Low (major unknowns)	Ask 1-2 targeted questions, then draft

Draft first, ask with the draft (not before):

Section	Draft Autonomously	Flag With Draft
Abstract	Yes	"Framed contribution as X—adjust if needed"
Introduction	Yes	"Emphasized problem Y—correct if wrong"
Methods	Yes	"Included details A, B, C—add missing pieces"
Experiments	Yes	"Highlighted results 1, 2, 3—reorder if needed"
Related Work	Yes	"Cited papers X, Y, Z—add any I missed"

Only block for input when:

Target venue is unclear (affects page limits, framing)
Multiple contradictory framings seem equally valid
Results seem incomplete or inconsistent
Explicit request to review before continuing

Don't block for:

Word choice decisions
Section ordering
Which specific results to show (make a choice, flag it)
Citation completeness (draft with what you find, note gaps)

默认策略：保持主动。先交付草稿，然后持续迭代。

置信度水平	行动
高（代码仓库清晰、贡献明确）	撰写完整草稿并交付，然后根据反馈迭代
中（存在部分歧义）	撰写草稿并标出不确定项，继续推进
低（存在大量未知信息）	提出 1～2 个有针对性的问题，然后开始撰写

先起草，并随草稿一起提问，而不是在起草前提问：

章节	自主起草	随草稿标注
摘要	是	“将贡献定位为 X——如有需要请调整”
引言	是	“突出问题 Y——如果判断有误请纠正”
方法	是	“包含细节 A、B、C——可补充缺失内容”
实验	是	“突出结果 1、2、3——可按需重新排序”
相关工作	是	“引用了论文 X、Y、Z——可补充我遗漏的文献”

仅在以下情况阻塞并等待输入：

目标会议不明确（会影响页数限制和论文定位）
多种相互矛盾的论文定位看起来同样合理
结果似乎不完整或前后不一致
用户明确要求审阅后再继续

以下情况不要阻塞：

用词选择
章节顺序
具体展示哪些结果（自行选择并标注）
引用完整性（根据已找到的内容起草，并说明缺口）

The Narrative Principle

叙事原则

The single most critical insight: Your paper is not a collection of experiments—it's a story with one clear contribution supported by evidence.

Every successful ML paper centers on what Neel Nanda calls "the narrative": a short, rigorous, evidence-based technical story with a takeaway readers care about.

Three Pillars (must be crystal clear by end of introduction):

Pillar	Description	Example
The What	1-3 specific novel claims within cohesive theme	"We prove that X achieves Y under condition Z"
The Why	Rigorous empirical evidence supporting claims	Strong baselines, experiments distinguishing hypotheses
The So What	Why readers should care	Connection to recognized community problems

If you cannot state your contribution in one sentence, you don't yet have a paper.

最关键的洞察：论文不是实验的集合——它是一个由证据支撑、具有单一明确贡献的故事。

每篇成功的 ML 论文都围绕 Neel Nanda 所说的“the narrative”展开：这是一个简短、严谨、以证据为基础且读者关心的技术故事。

三大支柱（到引言结束时必须非常清晰）：

支柱	描述	示例
具体贡献是什么（The What）	在统一主题下提出 1～3 项具体的新主张	“我们证明 X 在条件 Z 下能够达到 Y”
为什么成立（The Why）	提供支持这些主张的严格实证证据	有力的基线，以及能够区分不同假设的实验
为什么重要（The So What）	解释读者为什么应该关注这项工作	与领域公认的重要问题建立联系

如果你无法用一句话说明贡献，那么论文尚未真正成形。

Paper Structure Workflow

论文结构工作流

Workflow 1: Writing a Complete Paper (Iterative)

工作流 1：撰写完整论文（迭代式）

Copy this checklist and track progress. Each step involves drafting → feedback → revision:

Paper Writing Progress:
- [ ] Step 1: Define the one-sentence contribution (with scientist)
- [ ] Step 2: Draft Figure 1 → get feedback → revise
- [ ] Step 3: Draft abstract → get feedback → revise
- [ ] Step 4: Draft introduction → get feedback → revise
- [ ] Step 5: Draft methods → get feedback → revise
- [ ] Step 6: Draft experiments → get feedback → revise
- [ ] Step 7: Draft related work → get feedback → revise
- [ ] Step 8: Draft limitations → get feedback → revise
- [ ] Step 9: Complete paper checklist (required)
- [ ] Step 10: Final review cycle and submission

Step 1: Define the One-Sentence Contribution

This step requires explicit confirmation from the scientist.

Before writing anything, articulate and verify:

What is the single thing your paper contributes?
What was not obvious or present before your work?

"I propose framing the contribution as: '[one sentence]'. Does this capture what you see as the main takeaway? Should we adjust the emphasis?"

Step 2: Draft Figure 1

Figure 1 deserves special attention—many readers skip directly to it.

Convey core idea, approach, or most compelling result
Use vector graphics (PDF/EPS for plots)
Write captions that stand alone without main text
Ensure readability in black-and-white (8% of men have color vision deficiency)

Step 3: Write Abstract (5-Sentence Formula)

From Sebastian Farquhar (DeepMind):

1. What you achieved: "We introduce...", "We prove...", "We demonstrate..."
2. Why this is hard and important
3. How you do it (with specialist keywords for discoverability)
4. What evidence you have
5. Your most remarkable number/result

Delete generic openings like "Large language models have achieved remarkable success..."

Step 4: Write Introduction (1-1.5 pages max)

Must include:

2-4 bullet contribution list (max 1-2 lines each in two-column format)
Clear problem statement
Brief approach overview
Methods should start by page 2-3 maximum

Step 5: Methods Section

Enable reimplementation:

Conceptual outline or pseudocode
All hyperparameters listed
Architectural details sufficient for reproduction
Present final design decisions; ablations go in experiments

Step 6: Experiments Section

For each experiment, explicitly state:

What claim it supports
How it connects to main contribution
Experimental setting (details in appendix)
What to observe: "the blue line shows X, which demonstrates Y"

Requirements:

Error bars with methodology (standard deviation vs standard error)
Hyperparameter search ranges
Compute infrastructure (GPU type, total hours)
Seed-setting methods

Step 7: Related Work

Organize methodologically, not paper-by-paper:

Good: "One line of work uses Floogledoodle's assumption [refs] whereas we use Doobersnoddle's assumption because..."

Bad: "Snap et al. introduced X while Crackle et al. introduced Y."

Cite generously—reviewers likely authored relevant papers.

Step 8: Limitations Section (REQUIRED)

All major conferences require this. Counter-intuitively, honesty helps:

Reviewers are instructed not to penalize honest limitation acknowledgment
Pre-empt criticisms by identifying weaknesses first
Explain why limitations don't undermine core claims

Step 9: Paper Checklist

NeurIPS, ICML, and ICLR all require paper checklists. See references/checklists.md.

复制以下检查清单并跟踪进度。每个步骤都包括起草 → 反馈 → 修改：

Paper Writing Progress:
- [ ] Step 1: Define the one-sentence contribution (with scientist)
- [ ] Step 2: Draft Figure 1 → get feedback → revise
- [ ] Step 3: Draft abstract → get feedback → revise
- [ ] Step 4: Draft introduction → get feedback → revise
- [ ] Step 5: Draft methods → get feedback → revise
- [ ] Step 6: Draft experiments → get feedback → revise
- [ ] Step 7: Draft related work → get feedback → revise
- [ ] Step 8: Draft limitations → get feedback → revise
- [ ] Step 9: Complete paper checklist (required)
- [ ] Step 10: Final review cycle and submission

步骤 1：定义一句话贡献

此步骤需要科学家明确确认。

开始撰写前，应明确表达并确认：

这篇论文的核心贡献究竟是什么？
哪些内容在研究开展前并不明显或并不存在？

“我建议将贡献定位为：‘[一句话总结]’。这是否准确概括了你认为最重要的结论？是否需要调整侧重点？”

步骤 2：起草图 1

图 1 需要特别关注——许多读者会直接查看该图。

传达核心思想、方法或最引人注目的结果
使用矢量图形（绘图采用 PDF/EPS）
图注应脱离正文也能独立理解
确保在黑白条件下仍清晰可读（8% 的男性存在色觉缺陷）

步骤 3：撰写摘要（5 句话公式）

来自 Sebastian Farquhar（DeepMind）：

1. What you achieved: "We introduce...", "We prove...", "We demonstrate..."
2. Why this is hard and important
3. How you do it (with specialist keywords for discoverability)
4. What evidence you have
5. Your most remarkable number/result

删除“Large language models have achieved remarkable success...”之类泛泛的开场句。

步骤 4：撰写引言（最多 1～1.5 页）

必须包括：

2～4 条贡献列表（双栏格式下每条最多 1～2 行）
清晰的问题陈述
简要的方法概述
方法部分最迟应从第 2～3 页开始

步骤 5：方法部分

确保他人能够复现：

提供概念性框架或伪代码
列出全部超参数
提供足以复现实验的架构细节
呈现最终设计决策；消融实验放在实验部分

步骤 6：实验部分

对于每项实验，明确说明：

它支持哪项主张
它与核心贡献有何联系
实验设置（详细信息放在附录中）
应观察到什么结果：“蓝线展示了 X，从而证明 Y”

要求：

误差棒及其计算方法（标准差或标准误）
超参数搜索范围
计算基础设施（GPU 类型、总计算时长）
随机种子设置方法

步骤 7：相关工作

按方法体系组织，而不是逐篇介绍：

良好：“一条研究路线采用 Floogledoodle 的假设 [refs]，而我们采用 Doobersnoddle 的假设，因为……”

不佳：“Snap 等人引入了 X，而 Crackle 等人引入了 Y。”

应广泛引用相关文献——审稿人很可能就是相关论文的作者。

步骤 8：局限性部分（必需）

所有主要会议都要求提供这一部分。出人意料的是，坦诚通常会发挥积极作用：

审稿人通常被要求不得因作者坦诚承认局限性而扣分
主动指出弱点，可以提前化解潜在批评
解释这些局限为何不会动摇核心主张

步骤 9：论文检查清单

NeurIPS、ICML 和 ICLR 都要求填写论文检查清单。参见 references/checklists.md。

Writing Philosophy for Top ML Conferences

顶级 ML 会议的写作理念

Sentence-level clarity, word choice, time allocation and what reviewers actually read are covered in references/writing-philosophy.md. Read it before polishing prose.

句子层面的清晰度、用词、时间分配以及审稿人实际会阅读的内容，详见 references/writing-philosophy.md。润色文字前应先阅读该文档。

Conference Requirements Quick Reference

会议要求快速参考

ML/AI Conferences

ML/AI 会议

Conference	Page Limit	Extra for Camera-Ready	Key Requirement
NeurIPS 2025	9 pages	+0	Mandatory checklist, lay summary for accepted
ICML 2026	8 pages	+1	Broader Impact Statement required
ICLR 2026	9 pages	+1	LLM disclosure required, reciprocal reviewing
ACL 2025	8 pages (long)	varies	Limitations section mandatory
AAAI 2026	7 pages	+1	Strict style file adherence
COLM 2025	9 pages	+1	Focus on language models

Systems Conferences (OSDI, NSDI, ASPLOS, SOSP): See the systems-paper-writing skill for page limits, templates, deadlines, and submission rules.

Universal Requirements:

Double-blind review (anonymize submissions)
References don't count toward page limit
Appendices unlimited but reviewers not required to read
LaTeX required for all venues

LaTeX Templates: See templates/ directory for all conference templates.

会议	页数限制	Camera-Ready 额外页数	关键要求
NeurIPS 2025	9 页	+0	强制要求检查清单；录用论文需提供通俗摘要
ICML 2026	8 页	+1	必须提供 Broader Impact Statement
ICLR 2026	9 页	+1	必须披露 LLM 使用情况；采用互审机制
ACL 2025	8 页（长文）	视情况而定	必须包含局限性部分
AAAI 2026	7 页	+1	必须严格遵循样式文件
COLM 2025	9 页	+1	聚焦语言模型

系统类会议（OSDI、NSDI、ASPLOS、SOSP）：有关页数限制、模板、截止日期和投稿规则，请参阅 systems-paper-writing skill。

通用要求：

双盲评审（投稿必须匿名）
参考文献不计入页数限制
附录篇幅不限，但审稿人没有义务阅读
所有会场均要求使用 LaTeX

**LaTeX 模板：**所有会议模板均位于 templates/ 目录中。

Using LaTeX Templates and Converting Between Venues

使用 LaTeX 模板及转换会议格式

Starting a paper from a bundled template (Workflow 4), compiling, template pitfalls, and converting between venues after a rejection (Workflow 3) are covered in references/latex-templates-and-conversion.md. Always copy the complete template directory from

templates/

rather than writing LaTeX preambles from memory.

如何从内置模板开始写论文（工作流 4）、编译论文、避免常见模板问题，以及被拒后转换会议格式（工作流 3），详见 references/latex-templates-and-conversion.md。始终应复制

templates/

中完整的模板目录，不要凭记忆编写 LaTeX preamble。

Citation Workflow (Hallucination Prevention)

引用工作流（防止幻觉）

⚠️ CRITICAL: AI-generated citations have ~40% error rate. Never write BibTeX from memory.

⚠️ 关键要求：AI 生成的引用约有 40% 的错误率。绝不要凭记忆编写 BibTeX。

The Golden Rule

黄金法则

IF you cannot programmatically fetch a citation:
    → Mark it as [CITATION NEEDED] or [PLACEHOLDER - VERIFY]
    → Tell the scientist explicitly
    → NEVER invent a plausible-sounding reference

IF you cannot programmatically fetch a citation:
    → Mark it as [CITATION NEEDED] or [PLACEHOLDER - VERIFY]
    → Tell the scientist explicitly
    → NEVER invent a plausible-sounding reference

Workflow 2: Adding Citations

工作流 2：添加引用

Citation Verification (MANDATORY for every citation):
- [ ] Step 1: Search using Exa MCP or Semantic Scholar API
- [ ] Step 2: Verify paper exists in 2+ sources (Semantic Scholar + arXiv/CrossRef)
- [ ] Step 3: Retrieve BibTeX via DOI (programmatically, not from memory)
- [ ] Step 4: Verify the claim you're citing actually appears in the paper
- [ ] Step 5: Add verified BibTeX to bibliography
- [ ] Step 6: If ANY step fails → mark as placeholder, inform scientist

Step 0: Use Exa MCP for Initial Search (Recommended)

If Exa MCP is installed, use it to find relevant papers:

Search: "RLHF language model alignment 2023"
Search: "sparse autoencoders interpretability"
Search: "attention mechanism transformers Vaswani"

Then verify each result with Semantic Scholar and fetch BibTeX via DOI.

Step 1: Search Semantic Scholar

python

from semanticscholar import SemanticScholar

sch = SemanticScholar()
results = sch.search_paper("attention mechanism transformers", limit=5)
for paper in results:
    print(f"{paper.title} - {paper.paperId}")
    print(f"  DOI: {paper.externalIds.get('DOI', 'N/A')}")

Step 2: Verify Existence

Confirm paper appears in at least two sources (Semantic Scholar + CrossRef/arXiv).

Step 3: Retrieve BibTeX via DOI

python

import requests

def doi_to_bibtex(doi: str) -> str:
    """Get verified BibTeX from DOI via CrossRef."""
    response = requests.get(
        f"https://doi.org/{doi}",
        headers={"Accept": "application/x-bibtex"}
    )
    response.raise_for_status()
    return response.text

Citation Verification (MANDATORY for every citation):
- [ ] Step 1: Search using Exa MCP or Semantic Scholar API
- [ ] Step 2: Verify paper exists in 2+ sources (Semantic Scholar + arXiv/CrossRef)
- [ ] Step 3: Retrieve BibTeX via DOI (programmatically, not from memory)
- [ ] Step 4: Verify the claim you're citing actually appears in the paper
- [ ] Step 5: Add verified BibTeX to bibliography
- [ ] Step 6: If ANY step fails → mark as placeholder, inform scientist

步骤 0：使用 Exa MCP 进行初始搜索（推荐）

如果已安装 Exa MCP，可使用它查找相关论文：

Search: "RLHF language model alignment 2023"
Search: "sparse autoencoders interpretability"
Search: "attention mechanism transformers Vaswani"

然后使用 Semantic Scholar 验证每条结果，并通过 DOI 获取 BibTeX。

步骤 1：搜索 Semantic Scholar

python

from semanticscholar import SemanticScholar

sch = SemanticScholar()
results = sch.search_paper("attention mechanism transformers", limit=5)
for paper in results:
    print(f"{paper.title} - {paper.paperId}")
    print(f"  DOI: {paper.externalIds.get('DOI', 'N/A')}")

步骤 2：验证论文存在

确认论文至少出现在两个来源中（Semantic Scholar + CrossRef/arXiv）。

步骤 3：通过 DOI 获取 BibTeX

python

import requests

def doi_to_bibtex(doi: str) -> str:
    """Get verified BibTeX from DOI via CrossRef."""
    response = requests.get(
        f"https://doi.org/{doi}",
        headers={"Accept": "application/x-bibtex"}
    )
    response.raise_for_status()
    return response.text

Example

bibtex = doi_to_bibtex("10.48550/arXiv.1706.03762") print(bibtex)


**Step 4: Verify Claims**

Before citing for a specific claim, access the paper and confirm the attributed claim actually appears.

**Step 5: Handle Failures Explicitly**

If you cannot verify a citation at ANY step:

```latex
% Option 1: Explicit placeholder
\cite{PLACEHOLDER_smith2023_verify}  % TODO: Could not verify - scientist must confirm

% Option 2: Note in text
... as shown in prior work [CITATION NEEDED - could not verify Smith et al. 2023].

Always inform the scientist:

"I could not verify the following citations and have marked them as placeholders:

Smith et al. 2023 on reward hacking - could not find in Semantic Scholar

Jones 2022 on scaling laws - found similar paper but different authors Please verify these before submission."

bibtex = doi_to_bibtex("10.48550/arXiv.1706.03762") print(bibtex)


**步骤 4：验证主张**

在针对某项具体主张引用论文前，应访问论文并确认所归属的主张确实出现在论文中。

**步骤 5：明确处理失败情况**

如果任何步骤都无法验证引用：
```latex
% Option 1: Explicit placeholder
\\cite{PLACEHOLDER_smith2023_verify}  % TODO: Could not verify - scientist must confirm

% Option 2: Note in text
... as shown in prior work [CITATION NEEDED - could not verify Smith et al. 2023].

始终告知科学家：

“我无法验证以下引用，并已将其标记为占位符：

Smith 等人 2023 年关于 reward hacking 的论文——未在 Semantic Scholar 中找到

Jones 等人 2022 年关于 scaling laws 的论文——找到了相似论文，但作者不同请在提交前验证这些引用。”

Summary: Citation Rules

摘要：引用规则

Situation	Action
Found paper, got DOI, fetched BibTeX	✅ Use the citation
Found paper, no DOI	✅ Use arXiv BibTeX or manual entry from paper
Paper exists but can't fetch BibTeX	⚠️ Mark placeholder, inform scientist
Uncertain if paper exists	❌ Mark `[CITATION NEEDED]` , inform scientist
"I think there's a paper about X"	❌ NEVER cite - search first or mark placeholder

🚨 NEVER generate BibTeX from memory—always fetch programmatically. 🚨

See references/citation-workflow.md for complete API documentation.

情况	行动
找到论文、获得 DOI 并获取 BibTeX	✅ 使用该引用
找到论文，但没有 DOI	✅ 使用 arXiv BibTeX，或根据论文手动录入
论文确实存在，但无法获取 BibTeX	⚠️ 标记为占位符并告知科学家
不确定论文是否存在	❌ 标记 `[CITATION NEEDED]` 并告知科学家
“我记得有一篇关于 X 的论文”	❌ 绝不能引用——先搜索，否则标记为占位符

🚨 绝不要凭记忆生成 BibTeX——始终以程序化方式获取。🚨

完整的 API 文档请参阅 references/citation-workflow.md。

Troubleshooting, Reviewer Criteria and Figures

常见问题、审稿标准与图表

Troubleshooting common writing problems, the criteria reviewers score on, and table/figure conventions are in references/issues-reviewers-figures.md.

常见写作问题的排查方法、审稿人评分标准以及表格和插图规范，详见 references/issues-reviewers-figures.md。

References & Resources

参考资料与资源

Reference Documents (Deep Dives)

参考文档（深入阅读）

Document	Contents
writing-guide.md	Gopen & Swan 7 principles, Ethan Perez micro-tips, word choice
citation-workflow.md	Citation APIs, Python code, BibTeX management
checklists.md	NeurIPS 16-item, ICML, ICLR, ACL requirements
reviewer-guidelines.md	Evaluation criteria, scoring, rebuttals
sources.md	Complete bibliography of all sources
from-research-repo.md	Workflow 0: turning a code repository into a paper
writing-philosophy.md	Time allocation, clarity, word choice, what reviewers read
latex-templates-and-conversion.md	Template setup, compilation, pitfalls, venue conversion
issues-reviewers-figures.md	Common issues, reviewer criteria, tables and figures

文档	内容
writing-guide.md	Gopen & Swan 的 7 项原则、Ethan Perez 的微技巧、用词
citation-workflow.md	引用 API、Python 代码、BibTeX 管理
checklists.md	NeurIPS 16 项检查清单，以及 ICML、ICLR、ACL 的要求
reviewer-guidelines.md	评估标准、评分、rebuttal
sources.md	所有来源的完整参考文献表
from-research-repo.md	工作流 0：将代码仓库转化为论文
writing-philosophy.md	时间分配、清晰度、用词、审稿人实际阅读的内容
latex-templates-and-conversion.md	模板设置、编译、常见问题、会议格式转换
issues-reviewers-figures.md	常见问题、审稿标准、表格与插图

LaTeX Templates

LaTeX 模板

Templates in

templates/

directory:

ML/AI: ICML 2026, ICLR 2026, NeurIPS 2025, ACL/EMNLP, AAAI 2026, COLM 2025
Systems (OSDI, NSDI, ASPLOS, SOSP): See systems-paper-writing skill

Compiling to PDF:

VS Code/Cursor: Install LaTeX Workshop extension + TeX Live → Save to auto-compile
Command line:
```
latexmk -pdf main.tex
```
or
```
pdflatex
```
+
```
bibtex
```
workflow
Online: Upload to Overleaf

See templates/README.md for detailed setup instructions.

templates/

目录中的模板：

ML/AI：ICML 2026、ICLR 2026、NeurIPS 2025、ACL/EMNLP、AAAI 2026、COLM 2025
系统类（OSDI、NSDI、ASPLOS、SOSP）：请参阅 systems-paper-writing skill

编译为 PDF：

VS Code/Cursor：安装 LaTeX Workshop 扩展和 TeX Live → 保存后自动编译
命令行：
```
latexmk -pdf main.tex
```
或
```
pdflatex
```
+
```
bibtex
```
工作流
在线：上传到 Overleaf

有关详细设置说明，请参阅 templates/README.md。

Key External Sources

主要外部来源

Writing Philosophy:

Neel Nanda: How to Write ML Papers - Narrative, "What/Why/So What"
Farquhar: How to Write ML Papers - 5-sentence abstract
Gopen & Swan: Science of Scientific Writing - 7 reader expectation principles
Lipton: Heuristics for Scientific Writing - Word choice
Perez: Easy Paper Writing Tips - Micro-level clarity

APIs: Semantic Scholar | CrossRef | arXiv

ML/AI Venues: NeurIPS | ICML | ICLR | ACL

Systems Venues: See the systems-paper-writing skill for OSDI, NSDI, ASPLOS, SOSP links and guides

写作理念：Neel Nanda: How to Write ML Papers - 叙事、“What/Why/So What”

Farquhar: How to Write ML Papers - 5 句话摘要公式

Gopen & Swan: Science of Scientific Writing - 读者预期的 7 项原则

Lipton: Heuristics for Scientific Writing - 用词

Perez: Easy Paper Writing Tips - 微观层面的清晰表达

API：Semantic Scholar | CrossRef | arXiv

ML/AI 会议：NeurIPS | ICML | ICLR | ACL

**系统类会议：**有关 OSDI、SOSP、ASPLOS、NSDI 的链接和指南，请参阅 systems-paper-writing skill

ml-paper-writing

Original

Translation

ML Paper Writing for Top AI Conferences

面向顶级 AI 会议的 ML 论文写作

Core Philosophy: Collaborative Writing

核心理念：协作式写作

⚠️ CRITICAL: Never Hallucinate Citations

⚠️ 关键要求：绝不能幻觉式生成引用

The Problem

问题

The Rule

规则

When You Can't Verify a Citation

无法验证引用时

Recommended: Install Exa MCP for Paper Search

推荐：安装 Exa MCP 以搜索论文

Workflow 0: Starting from a Research Repository

工作流 0：从研究代码仓库开始

When to Use This Skill

何时使用本 Skill

Balancing Proactivity and Collaboration

在主动性与协作之间取得平衡

The Narrative Principle

叙事原则

Paper Structure Workflow

论文结构工作流

Workflow 1: Writing a Complete Paper (Iterative)

工作流 1：撰写完整论文（迭代式）

Writing Philosophy for Top ML Conferences

顶级 ML 会议的写作理念

Conference Requirements Quick Reference

会议要求快速参考

ML/AI Conferences

ML/AI 会议

Using LaTeX Templates and Converting Between Venues

使用 LaTeX 模板及转换会议格式

Citation Workflow (Hallucination Prevention)

引用工作流（防止幻觉）

The Golden Rule

黄金法则

Workflow 2: Adding Citations

工作流 2：添加引用

Example

Example

Summary: Citation Rules

摘要：引用规则

Troubleshooting, Reviewer Criteria and Figures

常见问题、审稿标准与图表

References & Resources

参考资料与资源

Reference Documents (Deep Dives)

参考文档（深入阅读）

LaTeX Templates

LaTeX 模板

Key External Sources

主要外部来源

Related skills

相关 Skills