Deep Learning Monitor

I’m in Fortune’s 40 under 40: https://t.co/a4cZrJVjvJ

2019-07-11 Super Hot

Recent progress in AI have taught us an important fact: an intelligence as general as human intelligence applied to… https://t.co/JBeQLZ60uE

2019-07-17 Super Hot

View point invariance is an important inductive bias in how we perceive objects - here tested to the limit by a sma… https://t.co/BFXGQCSb6u

2019-07-07 Super Hot

Bank of England governor picked Alan Turing as the face of the new £50 bill, among an impressive list of British sc… https://t.co/X65hAlFWz2

2019-07-15 Super Hot

@DeepMindAI

In our new paper https://t.co/Rhm94rOuX5 we show that GANs can be harnessed for unsupervised representation learnin… https://t.co/kvSH1nzwKN

2019-07-08 Super Hot

GPT-2 learned that it should use spaces rather than tabs 🙈🔥 https://t.co/SVK5XtxBKi

2019-07-18 Super Hot

I'll be giving a tutorial on “Neuroevolution for Deep Reinforcement Learning Problems” @GECCO2019 this weekend. Loo… https://t.co/BzhmLQYskU

2019-07-10 Super Hot

Drones in Guiyang, China https://t.co/IHM3EbVg37

2019-07-17 Super Hot

@skirani I thought this was satire at first glance 🙈

2019-07-13 Super Hot

There is an old misconception that common sense is rooted in language. IMHO, common sense emerges from knowing how… https://t.co/81OdN4U7sk

2019-07-09 Super Hot

@AndrewYNg

My friend @erikbryn just stopped by the office. I always find his work on AI's impact on the economy enlightening;… https://t.co/LoaAw846Bq

2019-07-09 Super Hot

@AndrewYNg

Many companies outside the software industry are still in the early phases of adopting AI. Landing AI’s… https://t.co/6xTqO1rNFO

2019-07-12 Super Hot

@kchonyc

re "Why not consider other models? such as XLNet": I agree with the reviewer on the importance of time travel resea… https://t.co/XTt7O4AsCG

2019-07-12 Super Hot

Replace some layers of a BERT-like architecture with "product key memory layers" and get better perplexity for half… https://t.co/7InOMSVtWa

2019-07-12 Hot

Could machines understand this? https://t.co/4bzPpFH2rC

2019-07-13 Hot

Interesting point. It certainly resonates with my own experience. https://t.co/4NVDlV33Vu

2019-07-10 Hot

It’s interesting to see the pendulum swing back to representation learning. During my PhD, most of my collaborators… https://t.co/zyoJnBUEgD

2019-07-08 Hot

Recurrent Neural Processes is a generalization of Neural Processes to sequences by introducing a notion of latent t… https://t.co/K7x4dI4phw

2019-07-14 Hot

What does it mean to understand a neural network? https://t.co/zmeinOM8Vx https://t.co/BDgEHuAuwd https://t.co/ta8OgYjNTj

2019-07-16 Hot

Can your RL agent solve this level? https://t.co/FfDywF63q1

2019-07-16 Hot

@demishassabis

It was an honour and a joy to be at @sim_manchester today to celebrate Alan Turing, one of Britain’s greatest scien… https://t.co/6cDTOVOnYJ

2019-07-15 Hot

@karpathy

Large Scale Adversarial Representation Learning https://t.co/1ZffPc749V un/self-supervised learning is a highly fer… https://t.co/X0y4IqjWXl

2019-07-08 Hot

10-page (PDF) #DataScience Cheat Sheet covers concepts in Probability, #Statistics, Statistical Learning,… https://t.co/l2pIqaXl5j

2019-07-06 Hot

DVD-GAN https://t.co/3Ax26l77Rg

2019-07-16 Hot

I must be a 0.1x engineer

2019-07-13 Hot

@nsaphra

If you're trying to understand a complex computation graph of matrix information, try a test case where every dimen… https://t.co/7mUlpCcOAz

2019-07-15 Hot

Make people disappear easily https://t.co/R8EfFvMXns

2019-07-19 Hot

Awesome new paper from FAIR: 1. A new type of large-scale memory layer that uses product keys (FAISS-like indexing… https://t.co/5JMM6yzyAV

2019-07-12 Hot

@RemiMathis

– Vous faites quoi de vos vacances ? – On a commencé à entraîner un réseau de neurones, pour obtenir une intelligen… https://t.co/PXNlt5h3oz

2019-07-19 Hot

a 10x engineer https://t.co/eyOd994gri

2019-07-13 Hot

RoBERTa: FAIR's self-supervised NLP system is now top dog on the General Language Understanding Evlauation (GLUE) l… https://t.co/8kCOQSpLNg

2019-07-19 Hot

@DeepMindAI

We are excited to announce our work on multiagent evaluation appearing in @SciReports journal! This joint work with… https://t.co/7XX71e7Nla

2019-07-09 Hot

Research is a marathon, not a sprint. If it feels like a sprint, pick more ambitious goals. Overwork is indeed harm… https://t.co/aVNaf0pw2J

2019-07-11 Hot

@JoostNiemoller

Ja, daar hoor je dan weer niks over. Recente wetenschappelijke publicatie: 'Geen bewijs voor aandeel van de mens bi… https://t.co/KDx7xPGPHm

2019-07-13 Hot

Image Registration: From SIFT to Deep Learning. Useful to know. https://t.co/iL7V7SuQRl

2019-07-18 Hot

@BlackHC @PyTorch @TensorFlow @scikit_learn Sure but...45% accuracy is not exactly good. You can get close to 88% w… https://t.co/yJHeYBIpUM

2019-07-17 Hot

@demishassabis

Turing’s story reminds us that though we’ve made progress, we mustn’t stop until everyone regardless of background,… https://t.co/f5wEBgsC0Q

2019-07-15 Hot

That 10x thread started new internet memes and made the term “10x engineer” a controversial term, so in the end things turned out okay ...

2019-07-14 Hot

@NandoDF Hi Nando, I'm not sure I agree with this. I believe we should always publish findings that we want others… https://t.co/yIEcLqu9te

2019-07-11 Hot

Slides of my GECCO2019 talk Some takeaways REINFORCE (Section 6 of Williams, 1992) *is* an ES, and is very closel… https://t.co/iQeuioUWQw

2019-07-13 Hot

Invariant Risk Minimization is a beautiful paper that I'll need to study deeply. Main idea: “To learn invariances… https://t.co/xaqvgsXhag

2019-07-12 Hot

@yoavgo

oh wow seems like this boring public hyperparameter search is going to take a while. https://t.co/wc4HAKkNnM

2019-07-18 Hot

Top Cheat Sheets for #MachineLearning Engineers: https://t.co/IQE6zTrcmU ————————— #BigData #DataScience #AI… https://t.co/r4o2J0i1aw

2019-07-13 Hot

Smart bird. https://t.co/28OCfrqH91

2019-07-07 Hot

Interesting talks from scaling up neural networks, a 30 year journey of @JeffDean, to detecting illegal deforestati… https://t.co/k2o3oweTAj

2019-07-10 Hot

@soumithchintala

It's great to see more work on GANs focusing on downstream semi-supervised tasks, and away from pretty-picture comp… https://t.co/vqOukHh35c

2019-07-08 Hot

While we were writing the original GAN paper, my co-author @dwf tried to get something similar to BiGAN working for… https://t.co/yIpn4rjsQr

2019-07-08 Hot

Jeff Dean will give a talk at University of Tokyo on 7/11 https://t.co/eJ6eEdfXID https://t.co/MkULH23pzs https://t.co/MBB3LyWrRS

2019-07-06 Hot

“A city dweller traveling through the desert will be completely dependent on a nomad to keep him alive. So long as… https://t.co/9SHJmBTg1g

2019-07-07 Hot

Conditional Density Estimation with Neural Networks: Best Practices and Benchmarks This looks like a pretty useful… https://t.co/mtGbo9SUnS

2019-07-06 Hot

@InertialObservr Thanks, I hate it.

2019-07-13 Hot

@LakeBrenden 1. I don't think that anticipating the potential behavior of other drivers more reliably than humans i… https://t.co/lIIiltjX3G

2019-07-17 Hot

The expectation shouldn’t be there. Inundating arxiv with incremental work does not accelerate science and it annoy… https://t.co/Od9zij3Nwk

2019-07-11 Hot

Our environment not only shapes our intelligence, but is also part of it. https://t.co/u0jFschBCA

2019-07-10 Hot

@DeepMindAI

These results showcase the potential of GANs and other generative models in unsupervised learning, as explained in… https://t.co/TJEYAbOMaO

2019-07-08 Hot

@demishassabis

Turing’s treatment by the country he served was a tragedy - the new £50 note is a powerful symbol of the recognitio… https://t.co/ymAvKL6Mkc

2019-07-15 Hot

“The non-ML way to do it is to build up an understanding of the material from the ground up that is solid enough to… https://t.co/rW9vrzGfzg

2019-07-12 Hot

@kchonyc

Understanding implicit regularization in deep learning by analyzing trajectories of gradient descent https://t.co/S5gLm5eKZR via @Inoreader

2019-07-11 Hot

@SpirosMargaris

This #AI magically removes moving #objects from #videos https://t.co/EQR5gSlAAC #fintech… https://t.co/5B9uGyDSdp

2019-07-17 Hot

MIT Says Their New #AI Programming Language Makes AI More Accessible to Everyone (goes beyond #DeepLearning):… https://t.co/WcTOuUfBNe

2019-07-07 Hot

@yoavgo

Thanks @honnibal and @_inesmontani for organizing #spacyIRL and for inviting me, I had a great time. It was super… https://t.co/cFAh0uZo75

2019-07-06 Hot

Introduction to Machine Learning: https://t.co/a2OHhwVM6Z ————— #BigData #DataScience #AI #MachineLearning… https://t.co/9W9fJ2fEhY

2019-07-19 Hot

@seanmcarroll

Is peer review before publication a good idea? In physics it’s basically been abandoned. We submit papers to… https://t.co/OyoNdmNRLo

2019-07-08 Hot

Best reviewer rebuttal ever. https://t.co/FdiwtOsR9s

2019-07-13 Hot

Comprehensive Repository of #DataScience and #MachineLearning Resources for #DataScientists :… https://t.co/ogsFVocdRp

2019-07-15 Hot

How humans encode faces https://t.co/qqZVT5O4fs

2019-07-08 Hot

@RandomlyWalking

@jmhessel “Self-supervised” is a rebranding for “unsupervised” to avoid confusing people who ask Qs like “how can L… https://t.co/QKSW5rAaGz

2019-07-15 Hot

@aussiastronomer

Me: I feel like the only thing that has consistently grown throughout my science career is my skepticism of error b… https://t.co/nvDwX3nm44

2019-07-17 Hot

@nsaphra

I have been checking daily for a recording of @prfsanjeevarora's extremely interesting ICML workshop talk, "Is Opti… https://t.co/Qgvpa4yB6v

2019-07-08 Hot

@JapanGov

Meet "Ray," an #AI robot guide at Tokyo's Shimokitazawa Station that answers to train-related questions in 4 langua… https://t.co/gkjNlzG1df

2019-07-16 Hot

@SpirosMargaris

Building Better #DeepLearning Requires New Approaches Not Just Bigger #Data https://t.co/YFOvu3afQ1 #fintech… https://t.co/BsJh69YAvw

2019-07-08 Hot

Striving for Simplicity in Off-policy Deep Reinforcement Learning “We believe it is crucial to study the relative… https://t.co/P7yJmoxf8M

2019-07-11 Hot

Dynamics-Aware Unsupervised Discovery of Skills https://t.co/yiObgKtOMX https://t.co/FkB10HgCWH

2019-07-06 Hot

“Research cannot be detached from the environment in which it is performed. Being in an environment where the chall… https://t.co/AL9kf9orvV

2019-07-15 Hot

Japanese curry chain @CurryIchibanya (CoCo壱番屋) will open its first restaurant in India early next year. 🤔 https://t.co/5izCfxclsB

2019-07-09 Hot

@random_walker

Fortunately, most IoT traffic today is encrypted. Unfortunately, even encrypted traffic reveals a lot about your ac… https://t.co/niseZRpPYV

2019-07-17