karpathy/arxiv-sanity-lite

Health v2 · maintenance only

Flags: no_releases

Adoption not part of the score

1681 stars · 208 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

A lightweight web application that periodically polls the arXiv API for new papers and lets users tag papers of interest. It recommends similar papers per tag using SVMs over tfidf feature vectors of abstracts, with a Flask web UI and optional daily email digests.

Use cases

• track new arxiv papers in my research area
• get recommendations of papers similar to ones I tagged
• receive daily email digests of new machine learning papers
• search and filter recent arxiv papers by topic
• self-host a personal arxiv paper feed

When to choose

• you want a simple, self-hostable arxiv tracker with content-based recommendations
• you prefer lightweight tfidf/SVM over heavyweight LLM-based pipelines
• you want daily email digests of new papers matching your tags

When to avoid

• you need full-text recommendations beyond abstracts
• you need a multi-user production service with heavy traffic
• you want modern NLP embeddings instead of tfidf features

Facets

application · maturity maintenance

search-engine machine-learning web-framework email machine-learning web-development python self-hosted arxiv paper-recommendations tfidf svm flask recommendation-system academic-papers research natural-language-processing web-server

Member repositories

For agents

markdown · JSON · MCP: product_card(name="karpathy/arxiv-sanity-lite")