Getting a paper accepted

https://news.ycombinator.com/rss Hits: 3
Summary

In 2019, I submitted a paper that was rejected with review scores 2.5, 3, 3. One week later, I resubmitted it with minor changes, and it was accepted with scores 4, 4.5, 4.5. For context, that’s an almost unspeakably dramatic jump in scores, from “middling reject” to “strong accept.” This post shows exactly those changes. We’ll frame them in two parts: Polish page 1 for acceptance Use the remaining pages to avoid rejection Page 1 has four parts: title, abstract, Figure 1, and introduction. We’ll make them specific, memorable, clear, communicate value, and hook the reader. Reviewers mostly decide accept vs reject by page 1. So we optimize the judgment-before-scroll. Then, to make sure our paper isn’t rejected, we’ll do due diligence in the rest of it by including stuff like baselines, ablations, statistical significance, and human evaluation. The tweaks that get the paper accepted—unexpectedly, happily—also improve the actual science contribution. But if you’re tempted to be evil, read this footnote. The full rejected and accepted submissions are available for download at the end. Page 1 Is 80% of Your Paper A paper has five parts: Title Figure 1 Abstract Introduction Rest of the paper Spend equal time on each of these. — Me misquoting Jitendra Malik quoting Don Geman Around 80% of a paper’s perceived quality is established on page 1. The title, Figure 1, abstract, and half the introduction are all there. It’s like a book’s cover. Throughout this post, I’ll show the rejected and accepted versions of the paper I mentioned at the top with the dramatic score swing. Here are both page 1s: Top: Left: Rejected page 1. Bottom: Right: Accepted page 1. First, consider page 1’s first impression: Is the Figure 1 colorful and eye-catching? Is the title unexpected? Maybe it has one intriguing word? Are there any curious terms (bolded or italicized)? Is the introduction (hopefully not) full of citations? Choose A Specific Memorable Title Rejected: Visually Grounded Comparative Lan...

First seen: 2025-05-22 02:23

Last seen: 2025-05-22 04:23