Adithya @AdithyaSub86

Searching for the purpose of life Joined September 2017

Tweets

102
Followers

4
Following

194
Likes

53

Adithya @AdithyaSub86

7 years ago

I just published An Introduction to Bayesian Inference a mathematical venture — part 2 link.medium.com/iZP9pl5F8R

0 0 1 0 0

Adithya @AdithyaSub86

7 years ago

I just published An Introduction to Bayesian Inference a mathematical venture link.medium.com/MNLDARkozR

0 0 1 0 0

Our new paper "Learning End-to-end Autonomous Driving using Guided Auxiliary Supervision", outlines a Multi-Task Learning from Demonstration framework for end-to-end autonomous driving in urban environments. With @AdithyaSub86 and Anbumani Subramanian. arxiv.org/abs/1808.10393

0 1 0 0 0

Marc G. Bellemare @marcgbellemare

7 years ago

Very excited our lightweight RL framework is now available for everyone to use! w/ @pcastr @carlesgelada @Subho87

Google AI @GoogleAI

7 years ago

Very excited our lightweight RL framework is now available for everyone to use! w/ @pcastr @carlesgelada @Subho87

10 366 858 0 52

6 55 173 0 8

Ian Goodfellow @goodfellow_ian

7 years ago

Evaluating generative models by estimating a latent skill variable that explains their performance seems like a promising new research direction

Catherine Olsson @catherineols

7 years ago

Evaluating generative models by estimating a latent skill variable that explains their performance seems like a promising new research direction

3 83 328 0 34

2 99 373 0 27

Ian Goodfellow @goodfellow_ian

7 years ago

Train ImageNet in 18 minutes for just $40. fast.ai/2018/08/10/fas… By my former colleague @yaroslavvb

1 136 585 0 38

Devi Parikh @deviparikh

7 years ago

Code + pre-trained models for FAIR's winning entry to the VQA Challenge 2018 is now open sourced! github.com/facebookresear… @tinajiangyu @vivnat @endernewton Marcus Rohrbach @DhruvBatraDB

1 30 85 0 3

Dr. Buzz Aldrin @TheRealBuzz

7 years ago

I was on the Moon! #Apollo11 @NASA

4K 56K 231K 0 343

Feryal @FeryalMP

7 years ago

Great tutorial on Normalizing Flows by @ericjang11 : Part 1 blog.evjang.com/2018/01/nf1.ht… Part 2 blog.evjang.com/2018/01/nf2.ht… Jupyter Notebooks github.com/ericjang/norma…

Alex Mordvintsev @zzznah

7 years ago

Great tutorial on Normalizing Flows by @ericjang11 : Part 1 blog.evjang.com/2018/01/nf1.ht… Part 2 blog.evjang.com/2018/01/nf2.ht… Jupyter Notebooks github.com/ericjang/norma…

5 84 403 0 18

Download Video

1 17 56 0 5

Sergey Levine @svlevine

7 years ago

Deep RL, from images, conditioned on user-specified goals, trains in a few hours on real robots: sites.google.com/site/visualrlw… Key idea: let the robot "practice" goals that it proposes to itself to learn more about the world. w/ @Vitchyr Ashvin Nair, M. Dalal, S. Bahl, S. Lin

1 75 228 0 22

Vinod Grover @vinodg

7 years ago

Our paper "Diesel: DSL for linear algebra and neural net computations on GPUs" will appear at MAPL 2018 dl.acm.org/citation.cfm?i…

1 16 89 0 6

Julian Togelius @togelius

7 years ago

You know what's better than video games for testing your RL algorithms? Why, a framework where you can easily make new games and levels! You can now use General Video Game AI @gvgai with @OpenAI gym. In a first paper, we test deep RL algos on GVGAI games. arxiv.org/abs/1806.02448

4 81 187 0 12

Download Gif

Sergey Levine @svlevine

7 years ago

Few-shot learning problems can be ambiguous. Now MAML can handle the ambiguity by sampling multiple classifiers via a Bayesian formulation of meta-learning. We call it PLATIPUS: arxiv.org/abs/1806.02817 w/ @chelseabfinn and @imkelvinxu some ambiguous celeba task:

1 37 146 0 8

Download Image

Depth First Learning @DepthFirstLearn

7 years ago

Announcing DepthFirstLearning.com! We are building a repository of study guides targeting consequential papers. Check it out, learn something in-depth, and help us build the next one. @avitaloliver @suryabhupa @kumarkagrawal @cinjoncin

6 153 395 0 40

Jacob Eisenstein @jacobeisenstein

7 years ago

Complete draft of a new textbook for NLP: github.com/jacobeisenstei… Thanks to everyone who gave me edits and corrections! Stop me at #NAACL2018 and I'll buy you a beer or a beignet.

11 288 954 0 111

Arnaud Doucet @ArnaudDoucet1

7 years ago

Hamiltonian variational auto-encoder: arxiv.org/abs/1805.11328. Using results from SMC samplers, we provide guidelines for Salimans-Welling MCMC/Variational setup. Based on Hamiltonian importance sampling ideas, we then provide a data-informed normalizing flow. Tempering is key.

0 43 132 0 14

Sergey Levine @svlevine

7 years ago

Seems like really simple model-based RL methods can do just as well at convergence as model-free RL, if uncertainty is handled properly, with orders of magnitude fewer samples: sites.google.com/view/drl-in-a-… with Kurtland Chua, @RCalandra, Rowan McAllister

6 71 253 0 17

Sergey Levine @svlevine

7 years ago

Inverse RL + meta-learning: if you've inferred goals from observing other agents' actions in the past, you should be better at inferring new goals more quickly. Meta-learning seems like a natural fit for IRL: arxiv.org/abs/1805.12573 (w/ @imkelvinxu Ratner, Dragan, @chelseabfinn)

0 65 213 0 22

Ian Goodfellow @goodfellow_ian

7 years ago

This paper shows how to make adversarial examples with GANs. No need for a norm ball constraint. They look unperturbed to a human observer but break a model trained to resist large perturbations. arxiv.org/pdf/1805.07894…