Mohsen Naghipourfar

Interested in Deep Learning and Computational Biology

[Hani Goodarzi @genophoria]

The "Woman, Life, Freedom" uprising in Iran was just three years ago. U.S. liberals were vocal in their support... a mural went up in Clarion Alley; protests were announced across SF. Today, their silence is deafening. I promise you that our community will remember this.

[Michael Poli @MichaelPoli6]

[1/7] Introducing Evo 2, a new foundation model for biology.
🚀 Evo 2 is the largest-scale, fully open-source AI model ever released: 40 billion parameters, over 9 trillion tokens, and a 1 million context length. All the details are public: weights, data, training infrastructure, and inference infrastructure.

⚡Evo 2 is built on a new model architecture: convolutional multi-hybrids (StripedHyena 2). StripedHyena 2 excels at modeling byte-tokenized data, providing faster training and lower perplexity compared to both Transformers and previous-generation hybrids based on state-space models.

I am grateful for the team behind Evo 2—working with you was one of the proudest moments of my career (the core pretraining team was fewer than five people; you can just do things).

📚 Today, we release two papers (yes, plural), as well as weights, data, training, and inference codebases. Enjoy!

[Dave Burke @davey_burke]

Introducing Evo 2 from Arc Institute - an AI that can model and design the genetic code for all domains of life. It’s one of the largest-scale truly open source AI models for biology (and in fact more generally - most “open source” large language models are only “open weights”)

[Patrick Hsu @pdhsu]

AI provides a universal framework that leverages data and compute at scale to uncover higher-order patterns

Today, @arcinstitute in collaboration with @nvidia releases Evo 2—a fully open source biological foundation model trained on genomes spanning the entire tree of life 🧵

[Hani Goodarzi @genophoria]

We are excited to introduce LoRNA-SH, a frontier RNA foundation model that leverages long-read transcriptomics and long-context inputs to learn and predict the transcriptome architecture with base-pair resolution: https://t.co/8hfWJ54lHS

[Hani Goodarzi @genophoria]

Check out our latest research by @MNaghipourfar and others in our team on building language models of coding sequence. In this preprint, we introduce cdsFM, a suite of codon-resolution language models!👇
https://t.co/l1pFKi8PWx

[Bahar @BaharZirak]

Our paper is out!!!
https://t.co/YkqpBGMGeD
Congrats to all the authors and thank you @genophoria for the opportunity!

Here is a short summary:

[Mo Lotfollahi @mo_lotfollahi]

1/9 CPAis finally published. It can predict single-cell responses to combinatorial perturbation (drugs, CRISPR). This is a joint collab between @meta and @fabian_theis @HelmholtzMunich. Read the thread to understand the LEGO analogy! https://t.co/Ia87dfgdBF

[Fabian Theis @fabian_theis]

So excited to see our compositional perturbation autoencoder (CPA) paper out @MolSystBiol! This generative DL model leverages single cell perturbation (drug, CRISPR etc) studies to predict cellular response. Great collab w @MetaAI! https://t.co/sDDwCgJTVg
https://t.co/HUBvO5pCDF

[Bahar @BaharZirak]

Super excited to share the results of our highly collaborative project co-led by @MNaghipourfar, Ali Saberi, @DelaramPB, Amirhossein Zarezadeh and supervised by my incredible mentor @genophoria.

[Saeed Souzangar | سعید سوزنگر @souzangar]

یک نفر نوشته بود اینترنت مگه وصله که برای کنکور میخوایید قطع‌اش کنید؟
#فیلترنت

[Brian Roemmele @BrianRoemmele]

I said the 2020s will bring about a revolution in electronic medicine.

@Novocure Tumor Treating Fields uses alternating electric fields specifically tuned to target cancer cells.

Spectacular technology that will only improve with new AI.

This is an "Apple moment" in medicine.

[Vahid Online @Vahid]

#جادی پس از آزادی، مثل همیشه با صورت خندان!
امیرعماد میرمیرانی با قید وثیقه از زندان آزاد شد. ۲۲ آذر #مهسا_امینی

[Kevin Fischer @kevinafischer]

Tired of forgetting terminal commands every 10 seconds?

⚡pip install terminal-copilot⚡

GPT-3.5 powered terminal commands - probably the biggest productivity hack you can install today :)

[cheshm_abi @chawshin_83]

عرفان زارعی دانشجویی که در مقابل آذری جهرمی سخنگوی دولت ایستاده بود و شجاعانه حرف میزد، روز ۱۶ آذر در مقابل دانشگاه توسط نیروهای امنیتی جمهوری اسلامی ربوده شد.

[Mahvash @MahvashJebeli]

لینک زیر رو کمک کنید کامل بشه، دانشگاه‌هایی که مدرک رسمی رو ازتون نمی‌خوان. با توجه به افزایش ناگهانی هزینه آزادسازی مدرک، کمک بزرگی می‌تونه باشه:

https://t.co/YvfqFb1fok

[Elnaz @Naz_Eliii]

بیشتر از یک ساعته که جمعیت زیادی جلوی سلف ایستاده و محکم گفته می‌شه « فکر نکنین یه روزه، قرار ما هر روزه». زیباست.

#دانشگاه_شریف

[hardmaru @hardmaru]

Can your self-driving car do this?
https://t.co/sVUWEzD6ES

[Mo Lotfollahi @mo_lotfollahi]

More exciting news on this: our work has made the cover of @NatureBiotech. AI helps democratize to analyze single-cell data! It basically works like a google map navigator to navigate you through the map of cells to analyze your data!

[Lisa Sikkema @SikkemaLisa]

We’re thrilled to finally share the integrated Human Lung Cell Atlas (HLCA), a single-cell reference atlas of 444 individuals, and 2.2M cells! https://t.co/iFTqRDucCc @humancellatlas Here’s a tweetorial on how we built it, and, most importantly, how you can use it: (1/20)