There’s an inherent pressure within the dissemination of analysis. On one hand, science thrives on openness and communication. On the opposite, guaranteeing high-quality scientific work requires peer opinions which might be typically prolonged and closed. In 1991, physicist Paul Ginsparg created the arXiv repository to alleviate a few of that pressure. The thought is that researchers have a spot to add their preprint manuscripts earlier than they’re revealed in a journal. The preprints are free to all however haven’t undergone peer review (there’s some screening).
Nonetheless, arXiv doesn’t facilitate open, two-way dialogue. Now, two Stanford college students have developed an extension of arXiv that creates a centralized public sq., of types, for researchers to debate preprints. IEEE Spectrum spoke with one of many two, Rehaan Ahmad, in regards to the challenge.
Rehaan Ahmad
Rehaan Ahmad is the cofounder of alphaXiv, which he started as an undergraduate challenge whereas at Stanford University, alongside fellow pupil Raj Palleti.
How does alphaXiv work?
Rehaan Ahmad: You possibly can change the “arXiv” within the URL to “alphaXiv,” and it opens up the paper and there’s feedback and dialogue. You possibly can spotlight sections and depart in-line feedback. There’s additionally a extra basic house web page the place you possibly can see what papers different individuals are studying via the positioning. It finally ends up being a pleasant method to filter for what papers are attention-grabbing and what aren’t.
What motivated you to create the positioning?
Ahmad: My cocreator Raj Palleti and I have been undergrads at Stanford doing analysis in robotics and reinforcement learning. We figured lots of people would have questions on papers, like us. So I put collectively a little bit mock-up two or three years in the past. It was simply sitting on my pc for some time. After which a 12 months afterward I confirmed it to Raj, and he stated we have to make this a public web site. We considered it as a model of Stack Overflow for papers.
How tough was it to construct?
Ahmad: Surprisingly tough! Our background is in analysis, and one of many more durable classes for this challenge is that writing analysis code versus precise code that works are two various things. For analysis code, you write one thing as soon as, you place it on GitHub, nobody will use it—and in the event that they do, it’s their drawback to determine. However right here, the positioning has been round for a 12 months and a half, and solely just lately have loads of the bugs been type of hashed out. The challenge began out on a single AWS server, and anytime somebody would submit about it, it might go viral, and the server would go down.
How do you hope alphaXiv can be used?
Ahmad: I see alphaXiv as simply connecting the world of analysis in a method that’s extra productive than Twitter [now X]. Folks discover errors in papers right here; individuals will learn their opinions. I’ve been seeing extra productive discussions with the authors.
Your advisors embody Udacity cofounder Sebastian Thrun and Meta’s chief AI scientist, Yann LeCun. How have your advisors contributed?
Ahmad: After the primary few months of working alphaXiv, we circulated rather a lot inside the computer-science neighborhood. However after discussing the platform with [University of Maryland physics professor] Victor Galitski, we realized having his voice and opinion to information choices that have been related to the physics neighborhood can be extremely vital. These concerned about computer-science papers are often extra within the trending/likes/filtering side of our web site, whereas these concerned about physics are often extra discussion-oriented.
This text seems within the April 2025 subject as “5 Questions for Rehaan Ahmad.”
From Your Web site Articles
Associated Articles Across the Internet
