Publishing on arxiv.org (or any other paper archiving services) has become an important part of research practice. Although using the service sounds straightforward, there are a few things to keep in mind, which is the topic I would like to discuss today.
Before we move on to the main topic, here are good reasons why we (researchers) want to use it.
- Putting up your paper on arXiv is fast and easy. If your manuscript is ready, you can just post it without any peer reviews, etc.
- Many people watch arXiv uploads; it’s a quick and easy way to advertise the work.
- One can claim the novelty by using the timestamps on the arXiv postings. If you have a good idea and you already have a manuscript, you can claim that you are the first one to publish the idea, instead of waiting for your paper to be accepted.
- It’s free!
Meanwhile, there are a few things you want to be careful about, especially if you are new to research publication practice.
Academic aspects
- arXiv articles are not peer-reviewed, so you can’t argue that it’s a serious piece of work. However, that doesn’t mean that it’s not a paper. It’s still public, and everyone can read it, considering that it’s your work. If the work is decent, people honestly don’t care where it’s accepted for official publication: they will start to cite the work. So, put a lot of effort into writing a good paper, as if you were writing for a conference.
- Your co-authors are going to have a say about your arXiv publication. Don’t post a random article with someone else’s name without them knowing about it. I will cover more later, but there could be some legal issues involved in here, too.
- Although you are confident about your work and you are itching to publicize it before anyone else claims the same idea, other people you work with can have a different opinion. To other people’s (more objective) eyes, your work could look premature, requiring more validation or proof. It’s not good for your own fame if you have a bunch of half-baked arXiv articles that no one reads. So try to work with others: they are going to act like the peer reviewers that the arXiv publication process is missing!
Ethical issues
- Once again, since it’s a publicized article, you cannot get away from all the usual ethics issues. Plagiarism is still a big deal here. Original articles are expected to be uploaded.
- LLM-hallucinations are still something to avoid. It’s your responsibility to prevent it from happening as an author.
- You shouldn’t ignore other people’s contributions. You should not add co-authors without their consent, either.
Legal issues
- Since it will be public and difficult to withdraw once uploaded, be careful about the copyright issues. Always choose the right license options for your articles. There are a bunch of different choices there, ranging from a completely free-to-use-and-modify one to more strict ones. Always consult with the senior co-authors, especially the person in charge (principal investigator, manager, etc.). If you are not sure which license to choose, do some studies. It’s good to know what they mean.
- Especially if company people are involved in the work, get their permission first. Companies might have complicated internal pre-approval processes before an article can be published. Your co-authors can be in an awkward situation if they find an unapproved article on arXiv.
- The internal process may also require a patent application, which is their mechanism for protecting the idea before it is published. Once again, talk to the co-authors before putting it up on arXiv.
- If the paper is eventually accepted and published at a conference or journal, you want to make it clear on your arXiv version that there is an official publication. That’s actually what most venues require.
Enough about the backgrounds. Here are a few good practices I can think of.
arXiv good practices
- Work on a research project as usual. Work on a paper deadline. Once the paper is ready, regardless of the venue’s review and publication schedule, you can consider posting it on arXiv.
- Be careful: some conferences and journals may impose restrictions on arXiv postings, e.g., an embargo period. Check on those policies before you post them.
- Always ask all the co-authors and discuss the arXiv posting plan. They might have a different idea or a different set of legal processes to follow.
- Don’t choose a random/default license option. Talk to your co-authors about it. If they don’t care, talk to an LLM, at least, to understand what it means to you.
- Always try to post something nice, instead of pushing half-baked ideas out. arXiv articles don’t help much in the job market or Ph.D. applications anyway. I recommend you be a little conservative here; it won’t hurt you.
- You will receive an email with a passcode once it’s up. Share the passcode with the co-authors so they can claim the authorship.
- Don’t forget to manage your articles. If you have a new camera-ready version, upload it as the next version. You can’t fix typos once your paper is published in conference proceedings or a journal, but you can fix the arXiv version (by uploading a new version), which is probably the most popular PDF version of the same paper.