Blaz Skrlj

🦀

[Cameron R. Wolfe, Ph.D. @cwolferesearch]

The ChatGPT API was released yesterday and it costs 90% less than expected. Here’s five methods (and resources to learn about them) that are **probably** being used to enable this price reduction… 🧵[1/6]

[Blaz Skrlj @SkBlaz]

Really nicely done!
https://t.co/oIx5rAK6bv

[Ching-Lung Hsu @hiallen72]

Hippocampal spatial representations exhibit a hyperbolic geometry that expands with experience | Nature Neuroscience https://t.co/cYJsBqjqOz

[George Papamakarios @gpapamak]

New paper: Equivariant MuZero
https://t.co/yR8mQdtKzH

A version of MuZero that incorporates environment symmetries by design.

Great internship work by @andreeadeac22 👏👏👏

[Mark Tenenholtz @marktenenholtz]

We've made so much progress in the last 7-8 years on complex domains like text and images and basically none on simple sequences of numbers (i.e. time-series forecasting).

In fact, I think we've regressed since good information is so hard to find.

[vicki @vboykis]

GitHub just wrote an article about how they had to write their own search engine (in Rust, for performance reasons) and a new probabilistic data structure to reduce indexing time because ES and Lucene were blockers for them but sure big data is over 😅

[DJ @DuaneJRich]

One day I'll get to studying Normalizing Flows. When I do, I'll start here https://t.co/RDFbMYxUT9

[George E. Dahl @GeorgeEDahl]

We've just released the first version of our Deep Learning Tuning Playbook! This is our attempt to distill our process for actually getting good results with deep learning. We emphasize hyperparameter tuning since it has been a large pain point. https://t.co/PjeJVWeOzS

[Rust Trending @RustTrending]

TheAlgorithms / Rust: All Algorithms implemented in Rust ★13080 https://t.co/n1m3fpHHzL

[Fermat's Library @fermatslibrary]

23456789 is the largest prime number with consecutive increasing digits

[Guodong Zhang @Guodzh]

How to train very deep NNs without shortcuts, but still achieve competitive results on ImageNet?

Our ICLR paper gives a simple solution derived from kernel approx theory. We hope this could enable further research into deep models.
https://t.co/8CsC54hJok

[Martina Maritan @MartinaMaritan]

animation work in progress🤓 - went from sphere to actual 3D Whole-Cell model #unity

[Russell Kaplan @russelljkaplan]

Today I saw the impact that AlphaFold is having on speeding up drug discovery firsthand:

[Blaz Skrlj @SkBlaz]

@Petkomat @toplovodar @Val202

[Yanai Elazar @yanaiela]

@BlancheMinerva Here's a few I like for different reasons (not exhaustive):

Careful perturbations:
https://t.co/QssBiBRe4y
https://t.co/mGhqvPBiAE

Cool ML modeling:
https://t.co/SNHHPiwI4Q
https://t.co/Na6Pmh7bhp

Data collection:
https://t.co/SFUN9vQYod
https://t.co/rjHQ8Bnkz2

[Colin McCarthy @US_Stormwatch]

Tonga's Hunga Tonga volcano just had one of the most violent volcano eruptions ever captured on satellite.

[AK @_akhaliq]

GradMax: Growing Neural Networks using Gradient Information
abs: https://t.co/IyAuQffXcS

a new approach for growing neural networks that aims to maximize the gradient norm when growing

[Blaz Skrlj @SkBlaz]

@Petkomat @choldgraf

[Blaz Skrlj @SkBlaz]

@Petkomat https://t.co/7Q3XYeksyr