🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

excited to finally share on arxiv what we've known for a while now: All Embedding Models Learn The Same Thing embeddings from different models are SO similar that we can map between them based on structure alone. without *any* paired data feels like magic, but it's real:🧵

This is a really fucking great thread. Confirms a suspicion that many have had for a while in a quantitative way. TLDR; after a certain point, the type model doesn't matter if you have enough parameters

Our CSO, Prof. Pasha Baranov, discusses the limitations of RNA-Seq and the necessity of Ribo-seq for a complete understanding. 📹 Watch the full video on YouTube here: youtu.be/YxrLpdYEoJ8 ✉️ Contact us at [email protected] #HarnessingTranslatomics

Flash invariant point attention enables the generation of larger biomolecules. Olivia Viessmann

Oklart hur många svenska följare jag har, men likväl. Jag kommer vara i Almedalen 23-26 Juni för Googles räkning och prata om AlphaFold, Gemini och AGI i största allmänhet. Om någon vill ses, säg till! googlesessions.se/program

when people were working on BERT i always found these types of visualizations compelling. seeing the attention mechanism in action is so cool why are they not popular anymore? do our models have too many layers for us to understand now? or are attention maps just not useful?

Cursor 1.0 is out now! Cursor can now review your code, remember its mistakes, and work on dozens of tasks in the background.

Excited to unveil Boltz-2, our new model capable not only of predicting structures but also binding affinities! Boltz-2 is the first AI model to approach the performance of FEP simulations while being more than 1000x faster! All open-sourced under MIT license! A thread… 🤗🚀

These results are amazing: AF2 struggles to predict structures of non-ideal de novo proteins, but fine-tuning on 10k design models (no exp validation) improves this, AND the fine tuning *works better* if we only tune on an experimentally stable subset (6k) of the 10k models

happy 8th birthday of maybe the most influential paper of all time 🥳

What would our data landscapes look like if we could biochemically characterize 100s, 1000s, 10^n evolutionary sequences? Papers like this make strides towards that moonshot dream!! Such neat work from Margaux Pinney, Ph.D. lab! science.org/doi/10.1126/sc…

🚨 New paper 🚨 RNA modeling just got its own Gym! 🏋️ Introducing RNAGym, large-scale benchmarks for RNA fitness and structure prediction. 🧵 1/9

Congratulations to Yousuf A. Khan and co-authors on the latest work (rdcu.be/euBhY), which reveals a unique “side loading” mechanism for Sec18/NSF-driven SNARE disassembly—offering fresh insight into membrane fusion! I’m proud to have collaborated on this work.

This 4th of July, I’m reflecting on the freedoms that fuel scientific rigor and public health. From all of us at NIH, wishing you a safe and healthy Independence Day!

I have never seen a thin person drinking Diet Coke.