Glad to have contributed to this study on the computational side.. 😃 Thanks to implosion carving, we can now build devices that operate in the visible wavelength regime for classification tasks. Machine learning driven physical device design is pretty exciting.. news.mit.edu/2026/powerfu...

Equivariance is dead! 😢 Or is it? 😈 Genie 3 is out! Our latest protein design model achieves SoTA results for binder design and motif scaffolding, greatly improving on BindCraft and Proteina-Complexa. It does so using all-atom SE(3)-equivariance based on a branched polymer representation👇

claude is good at making posters through html.. feels like doing everything programmically is better 😅 used to love editing stuff on illustrator

La proteina’s autoencoder can be thought of as a inverse folder to some degree 🤔 Wonder if just giving Ca, along with a random partial latent acts as a inverse folder (true we get the all atom structure as well)

Consider submitting your work to the GenBio workshop at ICML 2026! We have an amazing set of speakers whose work spans generative and agentic AI. Deadline: May 1, 2026 genbio-workshop.github.io/2026/

❤️

noticing that backbone design models trained on AFDB 588K, when sampled through ODE results in poor designability (~25%). But through a proper noise schedule in SDE it goes to 90% (as reported in lit.) this felt counterintuitive, wonder if the reason is, AFDB being just 42% designable to begin with

this [1] is nice, easier than having to type the passphrase of a ssh key all the time.. made a simpler guide [2] inspired by it [1] - gist.github.com/arianvp/5f59... [2] - gist.github.com/ramithuh/a7b...

In 10 years we might have an EEG headset that decodes thoughts, and routes to claude 😆 Would be a true brain computer interface 😅

it always amazes me..

training a flow matching model for protein backbone design (N, Ca, C) intra residue geometry is somewhat easy to learn, but inter-residue (peptide bond length, Ca-C-N and C-N-Ca angles) doesn't consistently match ideal stereochemical params [1] [1] www.ebi.ac.uk/thornton-srv...

it always amazes me that after a mac update there's more free storage than what was before updating

sometimes asking claude to start an agent to figure out a bug is more effective than asking claude directly. The reason being, the prompt i give is not as descriptive as the prompt it generates for the agent

its always satisfying to figure out what causes a DDP training job to get suck

its interesting how MSA is very important for protein structure prediction to extract co-evolution signals, but in the small molecule design/conformer generation scenarios with the scale we are working, the model can "learn physics"

im reorganizing one of my old repo coded in JAX, I'm starting to question whether in the age of claude code, how important is knowing syntax, etc.. 😅 feels like ideas matter more than the implementation at this time, where coding is straightforward if you ask the right fundemental questions

Scared to use less popular vscode extentions etc after reading this

today i learnt that you can filter runs on @weightsbiases.bsky.social using the git commit id

starting to love the @marimo.io notebook's reactive cell execution. all the nodes in the DAG are on the same page!

ran 3 python codes, all 3 were using the same cpu core while 19 were free 🤔 i thought the os is supposed to handle this

First, @tattabio.bsky.social is now on Bluesky!💙 and second, we launched mult-sequence CoSearch on SeqHub!

@marimo.io's hover tooltips showing function signatures & docs is nice

I like how @anthropic.com is doing research into the downsides of AI use.. so that we can improve how we interact/get help www.anthropic.com/research/AI-...

building a website through @typst.app is kinda straightforward.. was trying out this template: github.com/vsheg/tufted, forum.typst.app/t/tufted-web... for eg: Below you can see an example post from their template: new.ramith.fyi/posts/2025-1...

does flash attention variants support arbitrary attention biases 🤔

Taken together, we hope our results highlight the current limitations of the self-consistency evaluation that is so commonly used in the field, and thus encourage to establish new/additional criteria, or at least be more aware of the downsides of AF & co metrics when assessing designs. 18/19

Limitations of the refolding pipeline for de novo protein design [new] Refolding pipelines eval: evo info can hurt seq-struct compat.

was hoping to see some reasoning traces on the math countdown task.. instead, diffusion LLM, started saying "exactly" too many times 🤨 Maybe its "reasoning" is during denoising (remasking + unveiling), rather than the autoregressive case where it explicitly says ("oops i got that wrong")

Trying to train a diffusion LLM with reasoning on a math task (without SFT), and this is what it spits out 😒😑

glad this is available on Youtube!

A few py2Dmol updates 🧬 py2dmol.solab.org Integration with AlphaFoldDB (will auto fetch results). Drag and drop results from AF3-server or ColabFold for interactive experience! (1/4)

Silly thread for a Saturday: some of the #HPC clusters I’ve worked on over the years. First up is Cielo, a Cray XE6 I worked on at LANL! Which might actually be the prettiest supercomputer I’ve worked on.

Installing alphafold3 on the DGX Spark was pretty easy, but I forgot that JAX preallocates gpu memory* (XLA_CLIENT_MEM_FRACTION=0.95) Forgot to change those settings, and the spark froze without me being able to ssh 😆 *In the spark, it’s unified mem Gotto test again either reduced reallocation

atomworks is a pretty cool GitHub repo!

Messy thinker? That's a superpower! Find panel coming soon in Muse helps you jump straight to your thought in a busy board.

Are you a Ph.D. student at @cmu.edu? Can you describe your research and its importance in under 3 minutes to a layperson? In 2025, 🥇 went to a @cmurobotics.bsky.social student. 🥈🥉 to @cmuengineering.bsky.social. Register to compete in the 2026 #3MTCMU competition by 2/4. cmu.is/2026-3MT-Compete

An application asked for the lifetime number of lines coded.. with gemini's help wrote a code to count net lines (.cpp or .py) an author contributed to a repo.. Thinking of polishing it up further: current version: gist.github.com/ramithuh/aa6...

What will you do today, knowing that you are one of the rarest forms of life to ever walk the Earth? Mark Nepo

dgx spark training speed impression while training a diffusion language model on Shakespeare: DGX Spark: ~1.3 it/s L40: ~3.5 it/s (same batch size used here)

I like the la-proteina architecture : - transformer with Pair-Biased Multi-Head Attention | Model | Layers | Parameters | +---------+------+------------+ | Encoder | 12 | ~130M | | Decoder | 12 | ~130M | | Denoiser| 14 | ~160M | +---------+------+-----------+

Three more AF3-like models for #ProteinFolding Pearl: arxiv.org/abs/2510.24670 ODesign: odesign1.github.io/static/pdfs/... OpenFold (published earlier this week) Find the full list at: #ProteinFolding #AlphafoldandCo bit.ly/biofold

realizing that understanding something on various levels of granularity is important..

How to organize your talk? I used to present like this, thinking that I was being "academic", "organized", and "professional". BUT, from the audience's viewpoints, this sucks. 😱 Look how far they need to hold a long-term context to just make sense of what you're saying!

Is 3D dragging you down? Wish you could instead use the 2D ColabFold representation for all your work? 🤓 Introducing: py2Dmol 🧬 (feedback, suggestions, requests are welcome)

What a powerful song youtu.be/huc7IL7b8S8?...

Underrated advice: Have one friend who tells you the truth. Not the comfortable truth. The uncomfortable truth. The truth that makes you better, not feel better. That's love.

when you cant miss out a dinner w friends, yet as a TA you have to answer some questions students have* 🥹😅 *for a HW that's due midnight