This project is a modified by daily-arXiv-ai-enhanced tool. It automatically fetches new papers from arXiv.org, uses a Large Language Model (LLM) to summarize them, and publishes the summaries to a GitHub Pages website.
- Daily Updates: Automatically crawls arXiv daily for new papers in your chosen categories (e.g., biology, computer science).
- AI-Powered Summaries: Utilizes an LLM of your choice (e.g., DeepSeek Chat) to generate concise summaries.
- Customizable: Easily configure the arXiv categories, LLM, and summary language to fit your needs.
- Automated Deployment: Publishes the summaries to a clean web interface hosted on GitHub Pages.
First, fork this repository to your own GitHub account.
Navigate to Settings > Secrets and variables > Actions in your forked repository.
Create the following encrypted secrets:
OPENAI_API_KEY: Your API key for the LLM service.OPENAI_BASE_URL: The base URL for the LLM service's API endpoint.
Create the following variables:
CATEGORIES: A comma-separated list of arXiv categories to track (e.g.,q-bio.GN,cs.CV).LANGUAGE: The target language for the summaries (e.g.,Chinese,English).MODEL_NAME: The identifier for the LLM you want to use (e.g.,deepseek-chat).EMAIL: The email address to use for Git commits.NAME: The name to use for Git commits.
Navigate to the Actions tab in your repository and select the arXiv-daily workflow.
- Manual Run: You can click
Run workflowto manually trigger the process and test your configuration. - Scheduled Run: By default, the action is scheduled to run automatically at 16:30 UTC every day. You can customize the schedule by editing the cron job in
.github/workflows/run.yml.Note: The default time is set to coincide with typically lower API costs for some services.
To change the appearance of the web page, you should modify the index.html file.