Artemy Bakulin

PhD student in Computer Science @WeizmannScience | @YosefLab | Computational Biology & Deep Learning

[Artemy Bakulin @art_b_ai]

@ChanwooKim_ Wrong link, I think. Other than that congratulations, the preview is really interesting.

[Artemy Bakulin @art_b_ai]

@aryehazan It is no longer possible in the world where all thoughts had been thought by Schmidhuber

[Artemy Bakulin @art_b_ai]

@francoisfleuret How about Epps-Pulley test as in LeJEPA?

[Artemy Bakulin @art_b_ai]

@chrisoffner3d And both gram and ibot are applied to the patches, so it might be assumed as obvious that their synergy makes sense.

[Artemy Bakulin @art_b_ai]

@chrisoffner3d Gram acts as a kind of regularization, keeping the outputs of the model close to the outputs of the previous epoch model, which in turn is a better model for local properties and slightly worse for global.

[Artemy Bakulin @art_b_ai]

@chrisoffner3d I think they specify in the paper that ibot and dino losses optimize two different objectives. One ensures good local properties of the features, another good global. And while the global ones keep improving during training, the local ones degrade.

[Artemy Bakulin @art_b_ai]

No selfie with LeCun this time, unfortunately – but a jungle encounter with this zen monke would do for now.

[Artemy Bakulin @art_b_ai]

Had the best time at #ICLR2025 in Singapore! It was such a pleasure to present my work at MLGenx and LMRL workshops.

[IdoAmitLab @IdoAmitLab]

@FlorianIngelfi1 @YosefLab CytoVI is truly transformative for the immune, cancer, and clinical fields

[Hani Goodarzi @genophoria]

pyPAGE is now officially out! New spin on old problem... https://t.co/CQim8WD03t

[Vanja (Ivan Kulakovskiy) @halfacrocodile]

Join the IBIS Challenge: an open competition focused on the computational prediction of transcription factor binding motifs. IBIS aims to advance state-of-the-art methods for Inferring Binding Specificities of human transcription factors from diverse experimental data. (1/12)

[Dmitry Penzar @dmitrypenzar]

So, the Stanford Ribonanza RNA Folding competition is closed, the private leaderboard is now available. Our team has secured the first place!
https://t.co/9zsuZ9iahZ

Soon we will publish a detailed description of our solution

[Artemy Bakulin @art_b_ai]

And here is a link to the source code
https://t.co/hIgD3qoEMk

[Artemy Bakulin @art_b_ai]

Hey, everybody. As a follow-up to our pyPAGE paper we made a bunch of tutorials on how to apply it in various scenarios to perform differential gene-set enrichment analysis.
https://t.co/0TUBl5Y7ct

[Artemy Bakulin @art_b_ai]

And of course, don’t hesitate to check our miniature case study of #depression performed on #singlecell RNA-seq data. Genuinely, hope that pyPAGE joins the arsenal of tools which will effectively combat this disorder.
https://t.co/9fb1j9eXVL