Veteran reviewer: 'safe' cockroach papers get accepted, reviewers should reward flawed insight
plopesresearch · x · 2026-10-22
A researcher with 5 years of CVPR/NeurIPS/ICLR reviewing experience notes that 'safe' papers — valid questions, sufficient experiments — are the most likely to be accepted, likening them to Bill Freeman's 'cockroach' slide: everywhere, not exciting, yet hard to kill. A CHI reviewer responds that reviewers should value exciting papers with fresh insights even with minor flaws, since most classics had plenty of them.
More from Fun
- Claude overthinking a simple question meme: 'only happens in high thinking mode' — inductionheads · 2026-10-22
- Vibecoding addiction: excited and depressed at the same time, asks Paras Chopra — paraschopra · 2026-10-22
- "I'm 3 months behind AI KOLs but 2 years ahead of the public — still so early" — KevinNaughtonJr · 2026-10-22
- Yacine: Astra is so smart I have to 'harness the shit out of it' to keep it on rails — yacineMTB · 2026-10-22
- Building breakable cities with Opus 5.5 — open-source and playable now — smith2008 · 2026-10-22
- Robotaxi Riders Giving Birth? Thread Exposes the Hidden Humans Behind 'Full Automation' — danielrock · 2026-10-22