# How does arXiv prepublishing work with ICLR's double-blind review? [R] — retained record

> Retained reddit record from reddit:MachineLearning.

Canonical: https://highsignal.app/data/records/c555fda33d33855a

[← sources](https://highsignal.app/data)


source: [reddit](https://highsignal.app/data/reddit) (reddit:MachineLearning)  published 2026-10-23 11:51 ingested 2026-10-24 02:37

[original source ↗](https://www.reddit.com/r/MachineLearning/comments/1xdbvrf/how_does_arxiv_prepublishing_work_with_iclrs/)

## retained text



```
Hey folks,

maybe a really dumb question from an inexperienced PhD student. I recently submitted to ICLR and spotted a bunch of papers that I was already familiar with because I had seen them on arXiv before.

Doesn't this break the double-blind review process? I mean, on the one hand, people pre-publishing their work want to establish priority as quickly as possible, but on the other hand, they still want their papers to be treated as valid submissions in a double-blind review system.

I generally avoid pre-publishing my work for this reason, although this can cause some trouble when it comes to novelty claims, because I'm worried about getting desk-rejected.

I mean, I can just Google the paper title or even parts of the abstract and immediately find the preprint. It seems quite frustrating...
```



document fetched 2026-10-24 10:30 · canonical https://reddit.com/r/MachineLearning/comments/1xdbvrf/how_does_arxiv_prepublishing_work_with_iclrs
