Hank

AI scientist @ Illumina

[MedARC @MedARC_AI]

Today, we're excited to announce our first research competition!

INTRODUCING NANOPATH: a framework and challenge to train the best pathology foundation model in just 1 hour!

A quick thread on why we made this challenge and how to participate!

[Illumina @illumina]

When the challenge is clinical adoption of genomics 🧬 — who you gonna call? @illumina Medical Affairs!

Meet our US–Canada team supporting reproductive health and genetic disease testing. From publishing research that informs payer guidelines, to educating healthcare providers, to helping laboratories make the clinical and economic case for genomic testing, this crew plays a critical role.

🔗 Read our spotlight on this team and how they’re working behind the scenes to bring genomics into mainstream care: https://t.co/i9KxFmDtHS

[Seth Howes @SethSHowes]

I sequenced my genome at home, on my kitchen table.

I wrote up exactly how I did it - the equipment, protocol, theory, and cost:

https://t.co/Nkjqaho2zm

[AilaunchX @Ai_Tech_tool]

ANDREJ KARPATHY COULD HAVE CHARGED $2,000 FOR THIS COURSE.

He put it on YouTube.

The full training stack. Tokenization. Neural network internals. Hallucinations. Tool use. Reinforcement learning. RLHF. DeepSeek. AlphaGo.

3 hours of the most comprehensive LLM education that exists anywhere at any price.

Not how to use the tools.

How the entire system was built from the ground up and why it behaves the way it does.

The engineers who understand this build things the ones who only use the tools cannot even conceive of.

The gap between those two groups is not 3 hours.

It is everything those 3 hours quietly unlock for the rest of your career.

[Nebius @nebiusai]

See how GPU-accelerated clustering, deep learning-based protein embeddings, and high-speed homology detection come together to enable AI-driven annotation of non-CRISPR archaeal defense mechanisms.

[Illumina @illumina]

Today, we unveiled PromoterAI, a groundbreaking algorithm that, for the first time at scale, accurately deciphers pathogenic regulatory genetic variants in the noncoding regions of the human genome. In a study published today in @ScienceMagazine, researchers report a 6% increase in diagnostic yield for rare disease by applying this technology to a large patient cohort—a major step toward improving the diagnostic rate for patients. 🔗 https://t.co/jxtubxG6Qa

[Brian Hie @BrianHie]

We trained a genomic language model on all observed evolution, which we are calling Evo 2.

The model achieves an unprecedented breadth in capabilities, enabling prediction and design tasks from molecular to genome scale and across all three domains of life.

[Brian Clarke @brianfclarke]

🧵 1/9 Delighted to announce that #DeepRVAT is out today in Nature Genetics! Here's a recap of our tweetorial from the preprint, with updates for some exciting new results. It's been a joy to work with @Holtkamp_Eva, as well as @gagneurlab and @OliverStegle.

[Oliver Stegle @OliverStegle]

Really pleased to see DeepRVAT published. A new method for rare variant genetics by @Holtkamp_Eva & @brianfclarke in collab with @gagneurlab. I like to think of DeepRVAT as one network on top of another, harnessing variant annotations from bags of models- AlphaMissense and alike.

[Josep Ferrer @iamjosepferrer]

The Transformer's decoder clearly explained 👇🏻

[Yann LeCun @ylecun]

I'm old enough to remember that in France in the 1960s and 1970s, far-right chauvinists were accusing Portugese immigrants of eating cats.
It's an old trope.

Today, making such an accusation publicly against an ethnic group violates laws against "incitement to racial hatred" in most of Europe.
You may get sued, get a fine, and get blocked from running for office.

Elon and other American 1st Amendment maximalists will scream "tyranny!"
But anti-hate speech laws are the way European countries have learned to protect themselves against fascist takeovers.
And they sure know what they're talking about.

[Nikolaos Koutsouleris @koutsouleris]

👉Excited to share our latest study published in Schizophrenia Bulletin! @sihsarie used EEG & ML to uncover unique brain signatures of schizophrenia, depression, and electrophysiological aging: https://t.co/JW8rNvTnTc
#EEG #Biomarkers #EphysAGE #Schizophrenia #Depression

[Open Life Science AI @OpenlifesciAI]

Last Week in Medical AI: Top Research Papers/Models
🏅(September 1 - September 7, 2024)

Medical LLM & Other Models :

- CancerLLM: Large Language Model in Cancer Domain
- MedUnA: Vision-Language Models for Medical Image
- Foundation Model for Robotic Endoscopic Surgery
- Med-MoE: MoE for Medical Vision-Language Models
- CanvOI: Foundation Model for Oncology
- UniUSNet: Ultrasound Disease Prediction
- DHIN: Decentralized Health Intelligence Network

Medical Benchmarks and Evaluations:

- TrialBench: Clinical Trial Datasets & Benchmark
- LLMs for Medical Q&A Evaluation
- MedFuzz: Exploring Robustness Medical LLMs
- MedS-Bench: Evaluating LLMs in Clinical Tasks
- DiversityMedQA: Assessing LLM Bias in Diagnosis
- LLM Performance in Gastroenterology

LLM Digital Twins:

- Digital Twins for Rare Gynecological Tumors
- DT-GPT: Digital Twins for Patient Health Forecasting

Medical LLM Applications:

- HIPPO: Explainable AI for Pathology
- LLMs vs Humans in CBT Therapy
- ASD-Chat: LLMs for Autistic Children
- LLMs for Mental Health
- LLMs for Postoperative Risk Prediction

Frameworks and Methodologies:

- Rx Strategist: LLM-based Prescription Verification
- Medical Confidence Elicitation
- Guardrails for Medical LLMs

👇🧵 1/25

#Medical #AI #LLMs #top #papers #research #ml

[Gabriel Peyré @gabrielpeyre]

The Polyak-Łojasiewicz inequality is known to imply the linear convergence of function values. This great paper https://t.co/NsvBrqO1gB provides details and connections to other inequalities. What seems to be less well-known is that it also implies the linear convergence of the iterates themselves. This was explained to me by @BarboniRaphael and the proof is adapted from https://t.co/GxoajqZGdh.

[David Slattery @daslattery]

What an inspiring day - esp. in face of our ongoing IT, email issues! The 7th Junior Scientist Day in Frankfurt - ~100 local researchers attending, over 40 posters and presentations from our ECRs + a fantatic keynote from @DurstewitzLab on Deep Dynamical Systems to kick us off

[DurstewitzLab @DurstewitzLab]

How to analyze comput. & dynamic mechanisms of RNNs?

1/2 Our #NeurIPS2023 spotlight on a highly efficient algo for locating all fixed points, cycles, and bifurcation manifolds in RNNs: https://t.co/qmINjJB40h
@NeurIPSConf
By brilliant @Zahra__Monfared, @lukasironman, Nic Göring

[Hank @HanCao9]

@OliverStegle I really enjoy the conference yesterday, thanks for organizing it

[Hank @HanCao9]

Nice food, nice talks and nice people

[Oliver Stegle @OliverStegle]

What a nice event brining the community together! AI / health / life / physics.

[Eleonora Russo @russo_eleon]

Great Perspective piece on dynamical system reconstructions from @DurstewitzLab, @GeorgiaKoppe, and @maxinthur!

[Yann LeCun @ylecun]

Some folks: "Open source AI must be outlawed."
Open source AI startup community in Paris: