sidphbot/Auto-Research

Orchestrates arXiv paper retrieval and NLP processing to extract structured content (abstracts, highlights, tables, images) from PDFs, then uses transformer models (BERT2BERT for titles, LED for abstractive summaries, sentence embeddings for clustering) to automatically organize papers into thematic sections with generated headings. Provides both CLI and Streamlit web interfaces, plus a Python API with modular research tools for document analysis, enabling fine-grained control over model selection and computational resource allocation across CPU/GPU configurations.

No commits in the last 6 months. Available on PyPI.