@zhengyiluo
Zhengyi “Zen” Luo
@zhengyiluo
9 months
So I was trying to cite ReLU and this is the first thing that pops up: , with 3000+ citations 😂 Folks this is NOT the ReLU paper! A number of popular, well-known paper has made this mistake it seems... (There is even this website "how to cite relu":…
Tweet media one
Tweet media two
7
8
69

Replies

@cHHillee
Horace He
@cHHillee
9 months
@zhengyiluo This came up on r/ml a while ago haha:
Tweet card media
From the MachineLearning community on Reddit: [D] Am I the only one finding this a bit upsetting?

Explore this post and more from the MachineLearning community

www.reddit.com
1
0
6
@zhengyiluo
Zhengyi “Zen” Luo
@zhengyiluo
9 months
@cHHillee 😂😂😂 it’s pretty misleading
0
0
0
@alixleroyguzman
Alix Leroy 🌱
@alixleroyguzman
9 months
@zhengyiluo So, which one of the two should we cite then?
3
0
1
@zhengyiluo
Zhengyi “Zen” Luo
@zhengyiluo
9 months
@alixleroyguzman Looks like most people in DL cite the second one from Hinton. I would probably just do both🤷
1
0
1
@rendope
randy
@rendope
9 months
@zhengyiluo ah yes, the piecewise linear function was invented in 2018. before that, the world functioned using only straight lines.
0
0
1
@TommyLofstedt
Tommy Löfstedt
@TommyLofstedt
9 months
@zhengyiluo I remember reading the Fukushima paper for the first time, and nodding approvingly when I realised he used ReLU ;-)
0
0
0
@amir_nlp
Amir Kargaran
@amir_nlp
9 months
@zhengyiluo and this is not word2vec paper.
Tweet media one
0
0
3
@lbertinetto
Luca Bertinetto 🇪🇺 🌐
@lbertinetto
9 months
@zhengyiluo I feel ReLU has a no-citation-needed status, same as batchnorm, SGD and the likes.
0
0
0