Artem Artemev

Check out our work "Memory Safe Computations with XLA compiler" at #NeurIPS2022 (with Yuze An, @dyedgreen, @markvanderwilk). The paper and PR can be found at openreview.net/pdf?id=2S_GtHB… and github.com/tensorflow/ten…. The poster is neurips.cc/virtual/2022/p…. Some details in short [1/8]

@krzysztof_rus @PatrickKidger @ezyang I'm not sure how Enzyme is going to help here, even with MLIR support. User still needs an interface, in some form, to autodiff.

@PatrickKidger @ezyang Enzyme AD adds autodiff to LLVM but I am not sure if it plays nicely with MLIR.

My read on Mojo is it's what you would do if Swift for Tensorflow failed and you were like "why did it fail" and concluded it's because no one likes Swift, so instead you do Python, and also MLIR is cool so backend to MLIR directly instead of TF

@markvanderwilk will be at the #NeurIPS2022 presenting the poster. If you are at #NeurIPS pop in and say hello. Thanks! [8/8]

We also applied eXLA to the language transformer model, and in the experiment we modified the sequence length which in turn controls the size of the self-attention block. Out of the box TF implementation fails with OOM with lengths more than 2k, and eXLA runs up to 7k. [7/8]

Check out our work "Memory Safe Computations with XLA compiler" at #NeurIPS2022 (with Yuze An, @dyedgreen, @markvanderwilk). The paper and PR can be found at openreview.net/pdf?id=2S_GtHB… and github.com/tensorflow/ten…. The poster is neurips.cc/virtual/2022/p…. Some details in short [1/8]

When working with a Gaussian process, have you ever wondered why Cholesky factorization failed, or a CG solve did not converge? Answer: it's because you've got redundant, overlapping data points. And that's just the starting point! On arXiv now! arxiv.org/abs/2210.07893

Wide Mean-Field Bayesian Neural Networks Ignore the Data. (arXiv:2202.11670v1 [cs.LG]) ift.tt/iyfvQew

I am still welcoming PhD applicants for 2022 at Imperial College London. We are a growing research group, with clear goals on what new abilities we want to develop in ML and neural networks. Topics: Invariances, neural arch search, (Bayesian) model selection, Gaussian processes.

We are organizing a small-scale, offline #NeurIPS2021 satellite event in Cambridge (UK) on the 8th of December. If you are interested in NeurIPS content and are in the neighborhood, this is your chance to connect with your local machine learning community neuripsmeetupcambridge.info

Join us to discuss Conjugate Gradient based GP approximations! We make training easier by automatically setting approximation parameters like CG tolerance using marginal likelihood bounds. Today 5pm (London) / 9am PDT. Long talk and poster available at icml.cc/virtual/2021/p….

Current Conjugate Gradient Gaussian Processes require manual tuning to trade off accuracy and speed. Existing guidelines can give suboptimal results, without clear warnings. Our method tunes automatically, runs fewer CG steps, and performs better: arxiv.org/abs/2102.08314 👇1/6

I'm looking forward to speaking tomorrow. I will share some thoughts on: - How Gaussian processes can help deep learning - Recent work on accurate GP inference - What makes a method "exact", and to what extent recent methods live up to this Link below if you want to join!

Tomorrow 10 Dec at 11am GMT I will speak at the Bayesian Deep Learning Meetup about **Bayesian Model Selection** and how it can help architecture search. In a short 20 minutes we will discuss why we (Bayesians ∪ Deep Learners) should care, and approaches from now and the past.

Come and chat with the authors of our paper: Doubly sparse variational gaussian processes! aistats2020.net/poster_940.html #AISTATS2020 @stefanos_ele @aptemav @NicolasDurrande @jameshensman @PROWLER_IO We are more friendly than we look in the video ;)

Come and work with me. I'm recruiting doctoral students (see hict.fi/node/190, DL Aug 17) and post-docs (see arno.solin.fi) @CSAalto

My #ICML2020 tutorial videos on "Machine Learning with Signal Processing" are now freely available: I: slideslive.com/38930502/machi… II: slideslive.com/38930503/machi… III: slideslive.com/38930504/machi… IV: slideslive.com/38930505/machi… Slides: arno.solin.fi