This paper: “We train a classifier to predict whether a paper should be rejected based solely on the visual appearance of the paper. Results show that our classifier can safely reject 50% of the bad papers while wrongly reject only 0.4% of good papers.” 
https://twitter.com/miles_brundage/status/1075931121810993152?s=21 …
"Deep Paper Gestalt," Jia-Bin Huang: https://arxiv.org/abs/1812.08775
Training a model to pick up on the aesthetic of a paper.
@BrundageBot doesn't pick up on the gestalt of papers, which it judges based on abstracts. So it probably won't like this one, which doesn't mention RL 
Показать эту ветку
We can probably train a better model using data from 
1/n 
