Wolfgang M. Pernice

Asst. Prof | Laboratory for Digital Biology (LDB) @ the Merritt Center | Columbia University | ML + Rare diseases | More info:

[Wolfgang M. Pernice @joungMax]

My CVPR highlights🧵- but first some shameless promo: our paper on Interventional Style Transfer (IST) for OOD generalization is out! We argue real-world applications of ML/AI require robust OOD generalization & propose strategies to assess + improve:

https://t.co/hfvZvVqMPx

[Wolfgang M. Pernice @joungMax]

@OfficialLoganK Strong!

[Wolfgang M. Pernice @joungMax]

Go work with *the* team in visual representation learning and help us push a super high impact open frontier (with major biomedical applications).

[Piotr Bojanowski @p_bojanowski]

🔥 PhD in the DINO team 🔥
We have a PhD program opening in our team. Don't hesitate to reach out to me or Max if interested!

[Wolfgang M. Pernice @joungMax]

@josiezayner Pretty sure this is NVIDIA responding to Arc's Proto, both trying to preemptively claim a new interface layer. Little real value added now, but we'll def need better orchestration across biomedical software tools in the future.

[Wolfgang M. Pernice @joungMax]

@davidrliu @PrimeMedicine Truly remarkable and inspiring.

[Wolfgang M. Pernice @joungMax]

Been saying for years that AR+moldynamics could allow us get a "feel" for how a molecule behaves i.e. unlocm a whole new way to understand molecular bio by plugging into the human ability to process tactile information. These guys have built it! Super cool, can't wait to try.

[Fabio Cortes @fabiojcortes]

Five years ago, if someone had said my final task at @EPFL would involve prototyping a UI for molecular dynamics in AR, I would've called them crazy.

And yet, here we are, pushing the limits of what a web browser can do!

Built with @threejs and UIKit by @pmndrs

[Wolfgang M. Pernice @joungMax]

Great guide on the art of writing a good paper - with side-by-side comparisons! Highly recommend.

[Max Forbes @maxforbes]

working on a post that's basically "how to get a paper accepted," using a case study one of my own that went from reject (2.5, 3, 3) to accept (4, 4.5, 4.5) with just one week of revisions

[Wolfgang M. Pernice @joungMax]

@ashleevance Equity

[Wolfgang M. Pernice @joungMax]

@elonmusk absolute nonsense

[Wolfgang M. Pernice @joungMax]

An incredible opportunity to join a world leading team and push the frontiers in unsupervised deep learning across many high-impact scientific domains! 🔬🛰️🧪🤖

[Piotr Bojanowski @p_bojanowski]

🔥Our team working on unsupervised deep learning is looking for a PhD intern for 2025. We are looking for deep learning / machine learning PhD students! If you are interested, you can apply here: https://t.co/U8dt7CXwFl
Don't hesitate to reach out!

[Rachel Tobac @RachelTobac]

LinkedIn is now using everyone's content to train their AI tool -- they just auto opted everyone in.
I recommend opting out now (AND that orgs put an end to auto opt-in, it's not cool)
Opt out steps: Settings and Privacy > Data Privacy > Data for Generative AI Improvement (OFF)

[Wolfgang M. Pernice @joungMax]

We re living in a simulation 👍

[Wolfgang M. Pernice @joungMax]

And the name of the judge who just ruled Google a monopoly is... 🥁

[Wolfgang M. Pernice @joungMax]

@gitaramdharry My deepest condolences Gita. I m so sorry for your loss.

[Wolfgang M. Pernice @joungMax]

@dgmacarthur @nickywhiffin @NatureMedicine Similar story for us with a recent @NatureMedicine paper by DeepMind with effectively the same approach to fortify OOD generalization on med. images we published at CVPR for microscopy images before. No citation 🙁 Surely not intentional but still sucks. Can editors help?

[Wolfgang M. Pernice @joungMax]

@seldo That's what I do. Typically I only practice a few intro sentences to get going. I do spend a fair bit of time designing each slide and flow though so that by the end, the story tells itself.

[Wolfgang M. Pernice @joungMax]

That's a wrap on @NVIDIAGTC for me. It was a celebration of the immense potential of AI to empower human thought and capability - nowhere in my view is this potential greater than in its application to biology. The strong representation (and Jensen's support) of #DigitalBiology / #AIxBioMed at GTC speaks to it. 

But what's also clear is that too many current efforts here stay entirely in hypothesis-land. There is no way around the hard work of wet-lab experimental validation. Aviv made the case for the lab-in-the-loop like only she can. But let's go further: let's not only have the lab, but the patients themselves in the loop. 

If you want to learn more / committed the terrible sin of missing my talk, check out my presentation on Digital Biology for #RareDisease in about a month on the GTC platform (or just dm me 😊).

[Wolfgang M. Pernice @joungMax]

Armed with famous fruit-salad metaphors, there is no one that can communicate the most cutting edge biomedical research to a room full of technologists at @NVIDIAGTC like Aviv Regev. #rolemodel

[Chris Gibson @RecursionChris]

Great talk by ⁦@genentech⁩’s Aviv Regev. Was disappointed not to see more biopharma folks at #GTC2024, but not surprised at all Aviv is here and speaking to standing room only!

[Wolfgang M. Pernice @joungMax]

@timisstuck @CellCellPress 🍿🍿🍿

[Wolfgang M. Pernice @joungMax]

@NVIDIAGTC", @CathieDWood invokes an old madman joke "Half of our advertisement works. We just don't know which one". How much worse are we in biomedicine?

The promise of #digitalbiology is to (1) measure this, and (2) bring the fraction that works to 100% 🚀