arXiv PDF Info

A Dorsal inference model for identifying arXiv preprints.

This model hashes your local PDF and fetches structured metadata (Title, Authors, Abstract, Categories, etc.) from DorsalHub.

Useful For

• Identifying arXiv PDFs
• Generating citations

Compatibility

• Python: 3.11, 3.12, 3.13, 3.14
• Dependencies: Requires dorsalhub.

Quick Start

Run the model directly against a local PDF file (downloads and installs if not already installed):

dorsal run dorsalhub/arxiv-pdf ./2405.06604v1.pdf

Output Formats & Exporting

You can export to a number of formats from the CLI:

Example: Export to BibTeX (bibtex) citation:

$ dorsal run dorsalhub/arxiv-pdf ./2405.06604v1.pdf --export=bibtex
@misc{2405_06604,
  title = {Explaining Text Similarity in Transformer Models},
  author = {Alexandros Vasileiou and Oliver Eberle},
  eprint = {2405.06604},
  archivePrefix = {arXiv},
  primaryClass = {cs.CL},
  url = {[https://arxiv.org/abs/2405.06604](https://arxiv.org/abs/2405.06604)},
  year = {2024}
}

Example: Export a RIS (ris) citation:

$ dorsal run dorsalhub/arxiv-pdf ./2405.06604v1.pdf --export=ris
TY  - PREP
T1  - Explaining Text Similarity in Transformer Models
AU  - Alexandros Vasileiou
AU  - Oliver Eberle
AB  - As Transformers have become state-of-the-art models for natural language...
PY  - 2024
KW  - cs.CL
KW  - cs.LG
UR  - [https://arxiv.org/abs/2405.06604](https://arxiv.org/abs/2405.06604)
M3  - 2405.06604
ER  - 

Example: Export to Markdown (md):

$ dorsal run dorsalhub/arxiv-pdf ./2405.06604v1.pdf --export=md
---
arxiv_id: 2405.06604
year: 2024
categories: [cs.CL, cs.LG]
---

# Explaining Text Similarity in Transformer Models

**Authors:** Alexandros Vasileiou, Oliver Eberle

## Abstract
As Transformers have become state-of-the-art models for natural language...

Development

Running Tests

We use pytest for unit testing.:

pip install -e ".[dev]"
pytest --cov=arxiv_pdf tests/

Or via uv:

uv run pytest

License

This project is open source and provided under the Apache 2.0 license.