Browse all
arXiv research papers — all
All arXiv papers (AI/ML focus) — filter and search by category and year.
2,300 results
Source: official U.S. government open data. This is an organized index, not an official U.S. government site. "Explained" links to our summary page; otherwise links go to the official primary source.
About this dataset
This list holds AI and machine-learning papers posted to arXiv, the preprint server. arXiv is where authors publish manuscripts themselves before peer review, and in this field it is where results typically appear first.
How to read it
- Preprints are not peer reviewedAppearing on arXiv does not mean anyone has verified the results. The field moves fast enough that preprints are cited widely, but their conclusions should still be read differently from peer-reviewed work.
- Versions get revisedThe same paper is updated as v1, v2, and so on. When citing or comparing, check which version you are looking at.
- Categories are author-declaredLabels such as cs.LG (machine learning) or cs.CL (natural language processing) are chosen by the submitter, so work on one topic can span several categories.
FAQ
What is arXiv?
An open repository of preprints in physics, mathematics, computer science, and related fields, operated by Cornell University.
Can preprints be trusted?
They have not been peer reviewed, so the reader has to judge the work. The same research is commonly published later through a peer-reviewed conference or journal.