We recently developed a new, unsupervised version of capsule networks (with @sabour_sara, @yeewhye, and @geoffreyhinton). I hope that my new blog will make it easier to understand some ideas that led to this work. Enjoy!https://t.co/c1MuHRKYOD pic.twitter.com/s22uL3pGlH
— Adam Kosiorek (@arkosiorek) July 22, 2019
Here's the link to their adversarial dataset for argument comprehension. Would've been nice (and made the paper more interesting) if they also provided more interesting adversarial examples than just one (about Google 😉), perhaps in the Appendix section. https://t.co/WUhuTR0gq6
— hardmaru (@hardmaru) July 22, 2019
Contrary to popular belief, training a gigantic model on a humongous dataset of human text will not lead to AGI. 🙈🧠
— hardmaru (@hardmaru) July 22, 2019
Probing Neural Network Comprehension of Natural Language Arguments: https://t.co/3AtozIV03O https://t.co/Y5zg0Dfgke
Neural Architecture Search is mainly applied to supervised learning recently, but many forget that the original NEAT paper on topology search was originally positioned as an RL method.
— hardmaru (@hardmaru) July 22, 2019
I expect to see a resurgence of architecture search methods for RL as they go hand-in-hand. 🐡 pic.twitter.com/a39W040w7u
A Fair Comparison Study of XLNet and BERT with Large Models: https://t.co/3GSKDMaXTg
— Russ Salakhutdinov (@rsalakhu) July 22, 2019
BERT's success in some benchmarks tests may be simply due to the exploitation of spurious statistical cues in the dataset. Without them it is no better then random. https://t.co/o3i8FtmC8z
— /MachineLearning (@slashML) July 21, 2019
How the Transformers broke NLP leaderboards https://t.co/iNf5qBccQd
— /MachineLearning (@slashML) July 20, 2019
Invariant Risk Minimization: an Information Theoretic View.
— Ferenc Huszár🇪🇺 (@fhuszar) July 19, 2019
My post on Arjovsky et al's latest paper, with a slightly different derivation of the IRM objective:https://t.co/APyOO1gyWx
RoBERTa: FAIR's self-supervised NLP system is now top dog on the General Language Understanding Evlauation (GLUE) leaderboard.https://t.co/YBrziEy5E5 https://t.co/YBrziEy5E5
— Yann LeCun (@ylecun) July 19, 2019
The new RoBERTa model from FAIR just edged out XLNet on the https://t.co/ryDQeo2HU2 nine-task leaderboard. Here's the information we have so far: pic.twitter.com/B9wPtR7bY5
— Sam Bowman (@sleepinyourhat) July 18, 2019
Just gave a talk on my recent line of work on ***Robust Deep Learning Under Distribution Shift*** at @SimonsInstitute. Topics incl. impossibilities, label shift, detection, importance weighting & domain-adversarial methods https://t.co/JrOLiV0jXX
— Zachary Lipton (@zacharylipton) July 18, 2019
Subspace Inference for Bayesian Deep Learning. In our new #UAI2019 paper we construct low dimensional subspaces for scalable Bayesian inference on modern deep nets (with code)! Mode connectivity makes a guest appearance. https://t.co/0NX7OzhLaB pic.twitter.com/OeFoegMrZX
— Andrew Gordon Wilson (@andrewgwils) July 18, 2019
