Xavier Sumba

Computer Scientist / music & sport lover. Applied Research AI Developer working with conversation AI. Interested in Bayesian Inference, NLP and GNNs. 🇪🇨✈️🇨🇦

[Bernhard Egger @VisionBernie]

I collected a list of resources for my future doctoral students @UniFAU at the @CogCoVi lab - on 1st of October, my first 3 doctoral students will start their adventures: here is the collection of all those resources that are hopefully helpful and might also help others?

1/n

[PyTorch Best Practices @PyTorchPractice]

Implementing RepVGG in PyTorch https://t.co/qmfwqkx41L

#deeplearning #machinelearning #ml #ai #neuralnetworks #datascience #PyTorch

[Xavier Sumba @xavier_sumba]

@SIGIReCom @SIGIRConf @mdr @jacopotagliabue @kallumadi was the workshop recorded? If so, can you plz share?

[Andrew Gordon Wilson @andrewgwils]

I'm extremely excited to officially announce our new library, GPyTorch, which has just gone beta! Scalable Gaussian processes in PyTorch, with strong GPU acceleration.
repo: https://t.co/5xr3DsIOoy
website: https://t.co/rXwlrpZ6UG

[Yi Tay @YiTayML]

Happy to share that we have updated and published v2 of our "efficient transformer" survey!

Major updates:
✅ Expanded our scope to sparse models and added a ton of new models!
✅ Wrote a retrospective post about the advances in the past year.

Link:https://t.co/nAaTLG8wOp

[Zachary Lipton @zacharylipton]

Tomorrow, we launch "The Art of the Paper", a course on the principles, mechanics, & culture of scientific writing. While core to our work, this material seldom gets a formal treatment. I'm excited, nervous, & grateful to @mldcmu for the creative freedom.

https://t.co/iHtTVV469X

[David Van Valen @davidvanvalen]

The ugly reality about diversity in the academy comes down to a simple question - who is allowed to be less than perfect? Far too often we demand that people who come from uncommon backgrounds to live a mistake free life just to be given a chance to demonstrate their gifts.

[Alfredo Canziani @alfcnz]

Insane!!! 🥳🥳🥳
Kevin's books drafts publicly released and figures' code open sourced!

[Petar Veličković @PetarV_93]

New 4-page position paper, in defence of message passing! 🕸️🌟

tl;dr: most (if not all) GNNs that go 'beyond MP' can be expressed as MP; you just have to modulate the input graph. It could be trivial, useful, or insightful -- but seems always possible.

https://t.co/cSFlg0GLDk

[Xavier Sumba @xavier_sumba]

@chaitjo I have been looking into drawing tools. Does anybody can contrast into the learning curve and advantages with illustrator?

[Pablo Samuel Castro @pcastr]

muy emocionante poder iniciar los esfuerzos por fortalecer el ecosistema de IA y machine learning en ecuador 🇪🇨
¡espero verlos a muchos compatriotas por ahí!

[Petar Veličković @PetarV_93]

Geometric & Graph ML were a 2021 highlight, with exciting fundamental research & high-profile applications.

@mmbronstein and I interviewed distinguished experts to review this progress & predict 2022 trends. It's our longest post yet! See 🧵 for summary.

https://t.co/9znOfOAXV7

[Rishabh (in SF) @rishabh16_]

🚨NEW WEEKEND READING🚨

I'm publishing "Graph Neural Networks for Novice Math Fanatics" – a primer on the math behind GNNs using colourful drawings & diagrams

https://t.co/6KT3B0y7QK

(I hope this becomes the definitive guide to GDL for those hoping to enter the field!!!)

[Emtiyaz Khan @EmtiyazKhan]

Tomorrow, I will give a tutorial on "ML from a Bayesian Perspective" at ACML 2021.
https://t.co/pyRhcB8hAz

I also wrote a summary for the slides, explaining why a Bayesian perspective is essential for machine learners https://t.co/xN4JxTSaD8

Slides: https://t.co/3Ht6wRJqtn

[Daniele Grattarola @riceasphait]

In our new paper, we introduce a unifying and modular framework for graph pooling: Select, Reduce, Connect.
We also propose a taxonomy of pooling and show why small-graph classification is not telling us the full story.

Arxiv: https://t.co/PeIEb4ZECh

Time for a 🧵on 🎱:

[Xavier Sumba @xavier_sumba]

@alexisjross Advices and scope of a research statement

[Petar Veličković @PetarV_93]

My compatriot showing how impossible is nothing when you're dedicated enough, no matter your background (see the entire thread). Truly inspiring!

Awesome job @gordic_aleksa! See you in the office soon. :)

[Yarin @yaringal]

This year's #BDL workshop at @NeurIPSConf will focus on the reliability of BDL in downstream tasks, with invited talks from practitioners and the two NeurIPS BDL challenges
https://t.co/3VdnGbpnVj
Please consider submitting extended abstracts by Oct 1, or posters by Dec 1

[Distill @distillpub]

Today Distill published two expository papers on Graph Neural Networks, listed below.

[Emtiyaz Khan @EmtiyazKhan]

The video of my talk at the “theory of continual learning” workshop (ICML 2020) is now available at slideslive. https://t.co/XBRwRh3Lde