ml-paper-writing

Write publication-ready ML papers for NeurIPS, ICML, ICLR, ACL, AAAI, and COLM with verified citations.

1|Updated Mar 5, 2026
One-click install
npx skills add https://github.com/Clay-HHK/claude-skills --skill ml-paper-writing-clay-hhk
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: ml-paper-writing
Source: https://github.com/Clay-HHK/claude-skills/tree/main/20-ml-paper-writing
Command: npx skills add https://github.com/Clay-HHK/claude-skills --skill ml-paper-writing-clay-hhk

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires semanticscholar, arxiv, habanero, requests, and includes references (resource) components.

What problem does it solve? Drafting ML conference papers from research repositories is slow and error-prone, and AI-generated citations have roughly a 40% error rate that risks desk rejection. This Skill structures the entire paper-writing workflow and enforces programmatic citation verification. ## Core Features & Use Cases - End-to-End Paper Drafting: Explores a research repository, identifies the contribution, and produces complete drafts of abstract, introduction, methods, experiments, related work, and limitations sections. - Hallucination-Free Citations: Searches Semantic Scholar and arXiv, verifies papers across multiple sources, and fetches BibTeX via DOI content negotiation instead of generating references from memory. - Conference Templates & Checklists: Ships LaTeX templates for NeurIPS 2025, ICML 2026, ICLR 2026, ACL, AAAI 2026, and COLM 2025, plus mandatory checklist guidance and format-conversion workflows for resubmission. - Use Case: Point the Skill at a research repo with experimental results, and it delivers a full first draft targeting ICLR with verified citations, flagging any references it could not confirm as explicit placeholders. ## Quick Start Use the ml-paper-writing skill to draft a NeurIPS submission from my research repository and verify every citation programmatically.

Frequently Asked Questions about ml-paper-writing

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I write an ML paper for NeurIPS or ICML from a research repository?▼

Start by exploring the repository to identify the main contribution and existing citations, then draft sections iteratively: Figure 1, abstract, introduction, methods, experiments, related work, and limitations. The Skill provides a step-by-step checklist and delivers a complete first draft for feedback.

How do I prevent AI-hallucinated citations in my paper?▼

Never generate BibTeX from memory. Search Semantic Scholar, verify the paper exists in at least two sources, and fetch BibTeX via DOI content negotiation through CrossRef. Unverifiable references must be marked as explicit placeholders for human confirmation.

What LaTeX templates are included for conference submissions?▼

Templates cover NeurIPS 2025, ICML 2026, ICLR 2026, ACL, AAAI 2026, and COLM 2025, each with official style files, bibliography styles, and example documents. Copy the entire template directory and verify it compiles before editing.

How do I convert a rejected paper to a different conference format?▼

Start fresh with the target template and migrate only content sections, never the LaTeX preamble. Adjust for page limits, add venue-specific requirements like ICML's Broader Impact Statement or ACL's Limitations section, and update references for blind review.

What are the page limits for major ML conferences?▼

NeurIPS allows 9 pages, ICML 8 pages, ICLR 9 pages, ACL 8 pages for long papers, AAAI 7 pages, and COLM 9 pages. References and appendices are unlimited at all venues, and most allow one extra page for camera-ready versions.

When must I disclose LLM usage in conference submissions?▼

ICLR requires disclosure when LLMs contributed significantly to research ideation or writing such that they could be considered contributors. Grammar checking and minor editing do not require disclosure. NeurIPS requires declaring LLMs used as core methodology components.