Any particular preference among all of them?

A nice snowy stroll in Oxford today

Nice, it is! Thanks!

Do you know if there’s an open-access copy available? I couldn’t find it outside of the paywall

‎Slopes: Ski & Snowboard ‎Take your snow days to the next level! Uncover detailed stats (and bragging rights) about your days skiing and snowboarding, ride with friends, log your memories, and replay your winter adventures to...

getslopes.com is likely one of the best apps I have in my phone.

It’s gorgeous, packs all the things I want, and an incredible tool when skiing with friends—I love beer+replays at the end of the day

Quite niche—only used when I’m skiing—thus a pleasant surprise when I rediscover it after summer

Curious what drove this floor. A big seller going online? A competitor? Also what drives this floor up? Is that a dip on the 6th a combination of Swifties that also attended NeurIPS.

Is there any trading firm already trading compute as a commodity? Do compute futures even make sense?

sfcompute.com

Thanks @workshopmlsb.bsky.social for letting us share our work!

🔗📄 bit.ly/plaid-proteins

Tbh I need to play more with that. But ideally I’d like it to also flag or follow up on LI, Bsky, etc. Like a personal relationship manager agent.

Next up - invited speaker!

Boltz-1: Democratizing Biomolecular Interaction Modeling

@jeremywohlwend.bsky.social

overview of results for PLAID!

1/🧬 Excited to share PLAID, our new approach for co-generating sequence and all-atom protein structures by sampling from the latent space of ESMFold. This requires only sequences during training, which unlocks more data and annotations:

bit.ly/plaid-proteins
🧵

I've got a load of random NeurIPS bade photos. Now I just need to wait for someone to push an AI agent to deal with follow-ups and such.

Same! Starter packs I guess?

AlphaGo - The Movie | Full award-winning documentary YouTube video by Google DeepMind

If you have not seen it, you really should.

m.youtube.com/watch?v=WXuK...

Next up - Invited Speaker!

Learning the determinants of protein folding energy landscapes from large-scale experiments

@grocklin.bsky.social

On the plus side, people are doing lots of cardio pilgrimaging from one side to the other.

So that’s good, especially after the chonks of meat served at the pre-MLSB reception.

The live stream on top is great! Rooting for you guys on the next iteration!

I completely get that. It’s a shame, hopefully the good reception and the amount of people attending this year—really good job on that btw—can help move the needle for the next one.

If it helps I can make loud noses and posts such that next year they are more optimally scheduled &/or located :)

Next up - Contributed Talk!

Controllable All-Atom Generation of Protein Sequence and Structure from Sequence-Only Inputs

@amyxlu.bsky.social Wilson Yang @kevinkaichuang.bsky.social Vladimir Gligorijevic @kyunghyuncho.bsky.social
Richard Bonneau Pieter Abbeel @ncfrey.bsky.social

Holi, MLSB is packed!

I see a poor grad student trying to figure out how to attach their poster to the ceiling.

ByteDance having a branch dedicated to drug discovery with a real intent to bring assets to the clinic

Why the East-West block divide between MLSB and AIDrugX? Couldn’t we have both next to each other? Or even better stagger the events?

Huh, I missed that one. Which session was that?

Very bold to do a live AI agent demo in a presentation.
Bolder to ask the agent to design a better binder on the spot and seeing it go to worse.
Boldest ask feedback from a med chemist in the audience on the result.

Massive kudos to the presenter, unfortunately he ran out of time for the last step.

The two sides of the NeurIPS workshop experience.

NLP-adjacent researcher: We’ve trained on all human knowledge ever recorded since the birth of Cleopatra.

Some material chemist working on catalysis: Best I can do is 6 samples.

Interesting theory, but I can see it. Chem/bio Trojan horses inside tech companies basically. Something about that feels pretty hopeful.

Do you know if that was the case here?

The bane of my existence.

Oh, I think I missed that last tech bro moment, but yes it’s a shame. The beginning of the talk was truly sweet and really great work, so it’s sad to see that it all went south.

Is that motivated by a clear business/market rationale (bio is the new tech, so better capture a part of the pie)? Are all founders copying David Shaw? Or is it just this gen’s fad: new rich kids buy a LVMH, new rich tech companies start drug discovery teams?

A speaker just started a talk by saying “I’m gonna skip the what are LLMs RAGs blah blah blah, ‘cus you know”

I think I’m gonna start using that template for all my future presentations