finiteattractor @finiteattractor

Software Engineer @Microsoft, Physics graduate, science enthusiast, stargazer... MindVentures.org Seattle Joined January 2011

Tweets

154
Followers

20
Following

147
Likes

146

Frank Nielsen @FrnkNlsn

4 years ago

Yet another nice book (undergraduate prerequisites only). The last chapter is the culmination of this book: It explains manifolds arising from group actions. link.springer.com/book/10.1007/9…

6 73 492 0 140

Download Image

Can NNs learn and enforce symmetries on their predictions? Inspired by Noether’s Thm, we do so by learning quantities that should be conserved. This method can recover known conservation laws in scientific data & can be applied to raw video data. Paper: lis.csail.mit.edu/noether

3 51 261 0 74

Download Image

Pedro A. Ortega @AdaptiveAgents

4 years ago

This was a serendipitous finding while working on a much larger project on uncertainty in control. Was rather surprising to find that such a simple rule can be equivalent to optimizing a free energy functional

Google DeepMind @GoogleDeepMind

4 years ago

3 53 326 0 65

Download Image

1 23 125 0 24

Cassie Kozyrkov @decisionleader

4 years ago

Fun with GANs this morning! Here's a high-level intro to #AI techniques for art: bit.ly/quaesita_drugs

6 8 109 0 10

Download Image

Rose Yu @yuqirose

4 years ago

New #ICLR2021 paper on Equivariant Continous Convolution for Trajectory Prediction. We leverage insights from #FluidDnamics to consider the internal symmetry in real-world trajectories. Paper: arxiv.org/abs/2010.11344 Code: github.com/Rose-STL-Lab/E… (1/3)

4 51 251 0 28

Download Image

Emma Bostian @EmmaBostian

5 years ago

I’ve been reading a really cool book called Weapons of Math Destruction which examines big data bias and inequality. High recommend! 📖

22 93 1K 0 201

Download Image

Steven Strogatz @stevenstrogatz

5 years ago

Our latest paper was published today. It’s about the classic (and amazing!) phenomenon of pendulum clocks that get in sync. Metronomes can do something similar, as you may have seen in videos on YouTube. Hopefully the intro of the paper is readable. Enjoy! aip.scitation.org/doi/10.1063/5.…

4 41 178 0 24

Joaquin Vanschoren @joavanschoren

5 years ago

Today, @iddo and me are presenting a tutorial on meta-learning at #AAAI2021. For those who can't make it, we made all materials (slides and videos) publicly available here: sites.google.com/mit.edu/aaai20… Happy to (try to) answer any questions you may have :). #AI #MachineLearning

3 47 220 0 62

Download Image

Kyle Cranmer @KyleCranmer

5 years ago

More cool stuff from @MilesCranmer 👇

Miles Cranmer @MilesCranmer

5 years ago

More cool stuff from @MilesCranmer 👇

5 61 363 0 76

Download Image

0 2 17 0 0

Steven Brunton @eigensteve

5 years ago

This is one of my favorite videos in the sparsity series: Sparse Regression with the LASSO We use LASSO all the time to discover interpretable models that prevent overfitting (like SINDy) Check out the YouTube video! youtu.be/GaXfqoLR_yI

5 53 321 0 72

Download Video

finiteattractor @finiteattractor

5 years ago

The iconic Arecibo Observatory last moments! youtu.be/gupwFSoZ4GQ

0 0 0 0 0

Microsoft Flight Simulator✈️ @MSFSofficial

5 years ago

We’ve partnered with @VirtualFly1 to give away their High-End flight controls + a Premium Deluxe copy of MSFS! Starting off with their awesome Yoko+ ‘The Yoke’! To enter: 1. Like us + @VirtualFly1 2. RT Winner will be announced Dec. 1st #MicrosoftFlightSimulator #virtualfly

70 1K 1K 0 4

Download Image

Hanie Sedghi @HanieSedghi

5 years ago

Exciting new result on understanding failure modes of OOD generalization

Vaishnavh Nagarajan @_vaishnavh

5 years ago

Exciting new result on understanding failure modes of OOD generalization

1 58 308 0 86

Download Image

0 7 81 0 6

prof-g @robertghrist

5 years ago

i learned about these from the wonderful paper of curto, degeratu, itskov, & morrison. this uses a piecewise-linear model for each neuron. everything is driven by the network interactions -- whether strongly or weakly inhibitory. 2/ arxiv.org/abs/1605.04463

1 1 12 0 3

Li Li @leeley18

5 years ago

We discover that the prior knowledge embedded in the physics computation itself acts as an implicit regularization that greatly improves generalization of machine learning models for physics. Please check out our recent paper: arxiv.org/abs/2009.08551

7 117 371 0 82

Download Image

Arash Vahdat @ArashVahdat

5 years ago

📢 NVAE source code is released! NVAE is deep hierarchical VAE w/ specially designed network architecture that can generate high-quality images & achieve SOTA log-likelihood. Happy coding! paper: arxiv.org/abs/2007.03898 code: github.com/NVlabs/NVAE w/ @jankautz at @NVIDIAAI

3 94 405 0 58

Download Image

Brandon Amos @brandondamos

5 years ago

In our new paper we scale model-based reinforcement learning to the gym humanoid by using short-horizon model rollouts followed by a learned model-free value estimate. Paper: arxiv.org/abs/2008.12775 Videos: sites.google.com/view/2020-svg With @sam_d_stanton @denisyarats @andrewgwils

3 62 248 0 29

Download Video

Vivek Verma @vcubingx

5 years ago

IT'S FINALLY OUT! After months of work, I'm happy to announce that the first video in a series on visualizing deep learning is out. It's on visualizing the structure of a neural network. RTs appreciated :) youtu.be/UOvPeC8WOt8

28 1K 4K 0 872

Download Image

Cliff Pickover @pickover

5 years ago

Robert Bosch has created a fascinating series of instances of the traveling salesman problem (TSP) that provide continuous-line drawings of well-known pieces of art. Source: bit.ly/2Cq3Pfh

2 17 69 0 3

Download Image

Massimo @Rainmaker1973

5 years ago

If you put some soap or ink on a leaf, you obtain a ink propelled boat, using the Marangoni effect. The ink or the detergent has a lower surface tension than water and its flow provides thrust to push the boat forward buff.ly/2tnyzW6 [gif: buff.ly/2X23epo]