Hai Son

[Elon Musk @elonmusk]

Even Discord has gone corpo …

[Lucy L. Gao @lucylgao]

"Double-dipping" - generating a hypothesis based on your data, and then testing the hypothesis on that same data - is dangerous. To see this, let's take data with no signal at all ... 1/

[Hanie Sedghi @HanieSedghi]

Have you been thinking about “what is being transferred in transfer learning?” and what parts of the network are in charge of that? We have some answers for you! https://t.co/74I8ec1Tix with @bneyshabur and Chiyuan Zhang

[Alexia Jolicoeur-Martineau @jm_alexia]

Funny to think that I had my work published in a serious academic conference with this table.

[Daniela Witten @daniela_witten]

The Bias-Variance Trade-Off & "DOUBLE DESCENT" 🧵

Remember the bias-variance trade-off? It says that models perform well for an "intermediate level of flexibility". You've seen the picture of the U-shape test error curve.

We try to hit the "sweet spot" of flexibility.

1/🧵

[Justin Johnson @jcjohnss]

Videos for my Fall 2019 course "Deep Learning for Computer Vision" are now on YouTube!

This is an evolution of @cs231n that I used to teach at Stanford:
- All content refreshed
- New topics: Transformers, Video, 3D, etc
- HW in @PyTorch + @GoogleColab

https://t.co/6nqZKTpmxv

[Matthias Niessner @MattNiessner]

All videos and slides for our Introduction to Deep Learning Lecture (I2DL) with @lealtaixe are now public!

Covering: ML basics, Neural Networks, Optimization, Training Processes, RNNs..

Videos: https://t.co/dQFQTiR5iP

Slides: https://t.co/GJ2Jn2GswA

#deeplearning @TU_Muenchen

[AK @_akhaliq]

NeRF in the Wild: Neural Radiance Fields for Unconstrained Photo Collections
pdf: https://t.co/KHINkovxPi
abs: https://t.co/XWSUJZeqMA
project page: https://t.co/7S682vbr18

[Jon Barron @jon_barron]

Several people asked me for advice on how to best set parameters for histogram thresholding, so I made a little interactive tool that lets you explore the parameter space. Playing with this was more fun than I expected! https://t.co/H5uBkmi3Bp

[Hidenori Tanaka @Hidenori8Tanaka]

Q. Can we find winning lottery tickets, or sparse trainable deep networks at initialization without ever looking at data?

A. Yes, by conserving "Synaptic Flow" via our new SynFlow algorithm.

co-led with Daniel Kunin
& @dyamins, @SuryaGanguli

paper: https://t.co/642Clw00me
1/

[Ting Chen @tingchenai]

Introducing SimCLR: a Simple framework for Contrastive Learning of Representations. SimCLR advances previous SOTA in self-supervised and semi-supervised learning on ImageNet by 7-10% (see next).

https://t.co/X5CXud0VwL

Joint work with @skornblith @mo_norouzi @geoffreyhinton.

[Hai Son @sonhai]

I use @happy_scribe to transcribe my audio and video files. It's super useful and easy to use. First 30 minutes are free: https://t.co/JtxdbFqHSF

[Hai Son @sonhai]

@Hung_Q_Ngo https://t.co/xh1teuDs Delegates are a concept from C# for storing member function pointers

[Hai Son @sonhai]

@daobachlien vì quyển sách đó edit như một chuyển khảo lịch sử, mà đã là nghiên cứu lịch sử ko ai ko tự mình, mấy tay phê là phê thế

[Hai Son @sonhai]

@daobachlien vi d/c Nguyen Vu khong nen cat tay minh tu tu the

[Hai Son @sonhai]

@nagyojang hi Chankyu, how r u?