Repolode
Submit
ChenLiu-1996

ChenLiu-1996/figures4papers

A collection of Python scripts for producing publication-quality figures in the style of top AI conferences and journals, plus an LLM skill for generating them.

View on GitHubWebsite
Stars
8,719
Forks
605
Open issues
1
Last commit
8 days ago

What is it?

Figures for Papers is a centralized collection of Python scripts written by Chen Liu to produce figures for academic publications. The repository organizes one folder per project and covers common chart types used in papers, including bar plots for quantitative comparison and composition breakdown, radar and line plots, concept diagrams, and trend plots.

The figures shown in the repository were published at venues such as Nature Machine Intelligence, ICML, NeurIPS, and ECCV. Some figures in the project were only partially made in Python and are included as reference material.

The repository also contains a scientific-figure-making skill, a folder of conventions and reference documents that AI coding agents such as Cursor, Claude Code, or Codex can follow to write consistent plotting scripts. It can be used by path reference or installed as a symlinked skill.

Key features

  • 01Python scripts for bar, radar, line, trend, and concept plots used in papers
  • 02Figure exports for publication, with a shared style and export helper convention
  • 03A reusable LLM skill with style rationale, API conventions, and reusable patterns
  • 04Usable by path reference in AI coding agents or installed as a symlinked skill

Use cases

  • Producing consistent, publication-ready charts for a machine learning or science paper
  • Learning how to structure matplotlib scripts for academic figure styling
  • Guiding an AI coding agent to generate figure scripts that match a shared style

Things to know

  • The license is listed as NOASSERTION, so the terms of reuse are unclear.
  • The README describes these as the author's own scripts for their papers, not a general-purpose library.

Topics

#matplotlib#scientific-visualization#figures#plotting#llm-skills#academic-writing#acl#cvpr#eccv#emnlp

FAQ

What is figures4papers?
A collection of Python scripts for producing publication-quality figures in the style of top AI conferences and journals, plus an LLM skill for generating them. Figures for Papers is a centralized collection of Python scripts written by Chen Liu to produce figures for academic publications. The repository organizes one folder per project and covers common chart types used in papers, including bar plots for quantitative comparison and composition breakdown, radar and line plots, concept diagrams, and trend plots.
Is figures4papers free to use? What license does it have?
figures4papers is open source under the Other license. Check the license terms before using it in commercial products.
Is figures4papers still maintained?
Actively maintained. The last commit to figures4papers was on Oct 12, 2026.
What can I use figures4papers for?
Producing consistent, publication-ready charts for a machine learning or science paper Learning how to structure matplotlib scripts for academic figure styling Guiding an AI coding agent to generate figure scripts that match a shared style
Where can I get figures4papers?
The source code and installation instructions are on GitHub: https://github.com/ChenLiu-1996/figures4papers

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects