I am personally writing code to recreate some ML research and I am getting very different results from what the paper describes. Is there a community that likes to vet this type of work to help divulge validity?
[link][9 comments]
submitted by MentalMasochist
[link][9 comments]