Daily Paper

This project automates the process of fetching, filtering, and summarizing research papers from ArXiv based on user-defined keywords. Using GitHub Actions and an external LLM for summarization and scoring, it generates daily markdown reports highlighting the most relevant papers.

Paper

• Daily Papers
• All Papers

How it works

1. Fetch: The pipeline fetches the latest papers from ArXiv using the ArXiv API.
2. Filter: The pipeline filters the papers based on user-defined keywords.
3. Summarize: The pipeline summarizes the papers using an external LLM.
4. Score: The pipeline scores the papers based on the summaries.
5. Highlight: The pipeline highlights the most important papers based on the scores.
6. Report: The pipeline generates a daily markdown report of the papers.
7. Commit, Push, and Pull Request: The pipeline automatically commits the report to the repository, pushes the changes, and creates a pull request.

Acknowledgements

This project draws inspiration from the following open-source projects:

• DailyArxiv

ChatGPT also helped in the development of this project.

License

This project is licensed under the MIT License - see the LICENSE file for details.