Robocrunch        AI

MathType    @MathType   ·   9/15/2021
Liouville's theorem states that every bounded function that is holomorphic in the whole complex plane must be constant. It is a classical result in #ComplexAnalysis. #MathType
 Reply      Retweet   29      Like     95    

Posted by:

  Similar Tweets  

Matthias Niessner    @MattNiessner   ·   7/16/2021
Common issue in deep learning projects is that complex method designs are adopted before basic debugging. Often this leads to situations where the data loader still loads black images, but we've already tried 10 loss function.... cuz a paper claimed these would improve results.
 Reply      Retweet   56      Like     596    

MathType    @MathType   ·   9/7/2021
Cyclotomic polynomials are polynomials whose complex roots are primitive roots of unity. They play an important role in algebraic #NumberTheory and #Galois theory, being used to build abelian field extensions. #MathType
 Reply      Retweet   3      Like     13    

Posted by:

MathType    @MathType   ·   7/16/2021
The linear transport equation models the variation of the concentration of a substance flowing at constant speed and direction. It's one of the easiest partial differential equations, and one of the few that admits an analytic solution. #PDE #Analysis #MathType
 Reply      Retweet   4      Like     25    

Posted by:

MathType    @MathType   ·   9/1/2021
The Gelfond-Schneider theorem is a partial positive answer to Hilbert's 7th Problem, giving a check for transcendability (not being the solution to ANY polynomial equation in rational coefficients) for a big class of numbers #Analysis #MathType
 Reply      Retweet   12      Like     62    

Posted by:

MathType    @MathType   ·   9/13/2021
Math fact: both the jacobian matrix and its determinant are referred to as “The Jacobian”. Confusing right? The Jacobian matrix of a vector-valued function, named after Carl Gustav Jacobi, is the matrix of all its first-order partial derivatives. #VectorCalculus #MathType
 Reply      Retweet   29      Like     116    

Posted by:

Khanh Nguyen    @khanhxuannguyen   ·   6/15/2021
“Enough” does not mean “efficient”. A two-layer neural network with sufficient width can approximate any function. But the width could grow exponentially with the complexity of the function. Deep nets are more efficient function appriximators.
 Reply      Retweet   1      Like     9    

Grzegorz Chrupała 🇪🇺    @gchrupala   ·   9/16/2021
If a function to read a CSV file needs to have 52 parameters, maybe that's an indication that something is not quite right.
 Reply      Retweet   1      Like     3    

LeighDrogen.eth    @LDrogen   ·   9/6/2021
My wife has been saying this for a long time, that the obesity is a function not of the diet but of what’s in the meat, it’s all the runoff and hormones If you go to Europe and eat the same stuff you lose weight Same with out in Montana
 Reply      Retweet   1      Like     5    

David Pfau    @pfau   ·   9/11/2021
And real progress in AI will probably not come until we have the equivalent of complex analysis for intelligence.
 Reply      Retweet        Like     13    

Oriol Vinyals    @OriolVinyalsML   ·   6/11/2021
MuZero removed simulators in MBRL vs AlphaGo. VQ Models for Planning generalize to partial observable & stochastic environments. How? 1. Discretize states w/ VQVAE 2. Train a LM over states 3. Plan w/ MCTS using the LM Led by @yazhe_li & @sherjilozair
 Reply      Retweet   47      Like     224    

John Carlos Baez    @johncarlosbaez   ·   7/8/2021
Cool fact: Setᵒᵖ is equivalent to a certain category of very nice boolean algebras. A 𝗯𝗼𝗼𝗹𝗲𝗮𝗻 𝗮𝗹𝗴𝗲𝗯𝗿𝗮 is a set with operations "and", "or", and "not" obeying all the usual laws of classical logic. Do you know how any set gives a boolean algebra? (2/n)
 Reply      Retweet   4      Like     36    

Paul Kedrosky    @pkedrosky   ·   8/24/2021
If we had CGM (continuous glucose monitors), but for inflammation markers and for lung function, I think we'd really be onto something. Someone invent that/those.
 Reply      Retweet   4      Like     83    

Olivier Grisel    @ogrisel   ·   6/10/2021
Interesting use case for advanced second order optimizers like Shampoo: distillation with function matching where overfitting never happens:
 Reply      Retweet   5      Like     22    

VentureBeat    @VentureBeat   ·   9/7/2021
Large language models aren't always more complex by @Kyle_L_Wiggers
 Reply      Retweet   2      Like     2    

  Relevant People  

Type and handwrite mathematical notation. Easily include quality math equations in your documents and digital content. Part of @wiris_science.
MathType 33.7

Matthias Niessner
Professor for Visual Computing & Artificial Intelligence @TU_Muenchen Co-Founder @synthesiaIO
Matthias Niessner 34.5

John Carlos Baez
I'm a mathematical physicist interested in saving the planet.
John Carlos Baez 45.3

Quanta Magazine
Big ideas in science and math. Because you want to know more. Launched by @SimonsFdn. Get our weekly newsletter:
Quanta Magazine 56.7

David Pfau
So far I have not found the science, but the numbers keep on circling me. Views are very much my own.
David Pfau 36.0

Tom Goldstein
Associate Professor at Maryland. I want my group to do theory and scientific computing but my students don’t listen to me so I guess we do deep learning 🤷‍♂️

Khanh Nguyen
PhD student in CLIP lab (@umdcs), researcher on NLP, deep learning and imitation/reinforcement learning.
Khanh Nguyen 17.8