Intermediate

arXiv Navigation

Master the tools and strategies for finding, tracking, and staying current with AI research papers.

What is arXiv?

arXiv (pronounced "archive") is a free, open-access repository where researchers post preprints - papers before formal peer review. Most AI research appears on arXiv first, often months before journal publication.

Key arXiv Categories for AI

CategoryCodeTopics
Machine Learningcs.LGGeneral ML methods, theory, optimization
Computer Visioncs.CVImage recognition, generation, video
NLPcs.CLLanguage models, translation, text understanding
AI Generalcs.AIReasoning, planning, agents
Stats MLstat.MLStatistical learning theory

Finding Papers

🔍

Semantic Scholar

semanticscholar.org - AI-powered search with citation graphs, TLDR summaries, and influence scores.

📖

Google Scholar

scholar.google.com - Broad search across all academic sources. Set up alerts for topics and authors.

📄

Papers With Code

paperswithcode.com - Links papers to their code implementations and benchmarks. Essential for reproducing results.

🔔

Connected Papers

connectedpapers.com - Visual citation graph. Enter a paper, see a map of related work. Great for literature surveys.

Staying Current

  1. Daily arXiv Digests

    Subscribe to arXiv RSS feeds for cs.LG and cs.CL. Use tools like arXiv Sanity or Hugging Face Daily Papers to filter the most relevant papers.

  2. Twitter/X and Bluesky

    Follow researchers who share and discuss papers. Many breakthroughs are first discussed on social media with plain-language explanations.

  3. Newsletter Digests

    Subscribe to The Batch (Andrew Ng), Import AI (Jack Clark), or AI Weekly for curated summaries.

  4. Conference Proceedings

    Track accepted papers at NeurIPS, ICML, ICLR, ACL, CVPR, and AAAI. Accepted papers have been peer-reviewed.

Understanding Citations

Follow the citation trail: When a paper cites another paper that sounds important, read that one too. Key foundational papers appear in many reference lists. Citation count (available on Semantic Scholar) is a rough proxy for influence, but new important papers start with zero citations.

arXiv Paper IDs

arXiv URL Format
# Standard URL format
https://arxiv.org/abs/1706.03762    # "Attention Is All You Need"
https://arxiv.org/abs/2005.14165    # GPT-3 paper
https://arxiv.org/abs/2302.13971    # LLaMA paper

# The ID format: YYMM.NNNNN
# 1706.03762 = 2017 June, paper #03762
# Download PDF: replace /abs/ with /pdf/
https://arxiv.org/pdf/1706.03762

Ready to Go Deeper?

Live instructor-led courses from our partners. Affiliate disclosure.