Welcome to NERxiv's documentation page.

Named Entity Recognition for arxiv papers (NERxiv) is a Python wrapper tool for extracting structured metadata from scientific papers on arXiv using LLMs and modern retrieval-augmented generation (RAG) techniques.

Tutorials

The Tutorials are designed to guide you through the general and basic understanding of the functionalities of NERxiv. These are intended for learning the software in a first approach.

• Using the RAG Extractor Agent

How-to guides

The How-to guides provide step-by-step instructions for a wide range of tasks. These are oriented for a practical application so you can use them as a fast entry point when searching for specific information.

• Create Custom Prompts
• Customize Chunking Strategies
• Configure Retrieval Models
• Use Different LLM Models
• Adjust LLM Parameters

Explanations

The Explanations give you a theoretical background about the main concepts used in NERxiv.

• What is RAG?
• Understanding Chunking Strategies
• How Semantic Retrieval Works
• Prompt Engineering for Metadata Extraction

References

The References include the glossary of used terms and API documentation.

• Glossary
• API Reference

Contact

Many parts of this documentation are currently under construction. If you have further questions that are not yet answered here, please contact jose.pizarro-blanco@bam.de.