#mLsky
LightNews LightNews
puhsu.bsky.social
#MLsky
November 29, 2024 at 5:33 AM
smcgrath.phd
🧪 🩺🖥️ #MLSky
February 4, 2025 at 5:17 AM
wesburdine.bsky.social
If petar musa scores against England, MLSky is going to be insufferable
June 17, 2026 at 7:34 PM
smcgrath.phd
Tagging 🧪 🩺🖥️ #MLSky
January 13, 2025 at 3:33 AM
eddiebullock.bsky.social
when can i buy parmesan cheese without having to take out a mortgage first.

#economy #bluesky #python #mlsky #cheese
August 7, 2025 at 4:55 PM
nsaphra.bsky.social
So is the “linear representation hypothesis” really just people noticing that nonlinear manifolds are locally linear? #mlsky
November 18, 2024 at 2:21 PM
tedunderwood.com
I find the #MLSky feed is quite good these days, for both papers and news.
February 12, 2025 at 2:21 AM
tyrellturing.bsky.social
Fantastic thread here, really worth a read!

#MLSky 🧪
karenhao.bsky.social Karen Hao @karenhao.bsky.social · Jan 27
As someone who has reported on AI for 7 years and covered China tech as well, I think the biggest lesson to be drawn from DeepSeek is the huge cracks it illustrates with the current dominant paradigm of AI development. A long thread. 1/
January 27, 2025 at 8:53 PM
eddiebullock.bsky.social
AI CEOs need to be more honest....

#science #AI #MLsky #google #meta #microsoft #openai #python #doomsday
August 12, 2025 at 4:16 PM
tedunderwood.com
I hadn’t even seen that they open-sourced these circuit-tracing tools. #MLSky
asura.dev asura @asura.dev · May 31
Is this in regards to Anthropic's tracing? Or just a tangent of language models?

It is a very wild ride to look through Gemma, etc
Open-sourcing circuit-tracing tools
Anthropic is an AI safety and research company that's working to build reliable, interpretable, and steerable AI systems.
www.anthropic.com
May 31, 2025 at 1:05 AM
eddiebullock.bsky.social
Do you journal?

#journal #research #mentalhealth #psychology #mlsky #dev #python
January 10, 2025 at 2:33 PM
puhsu.bsky.social
Hello to all #ICLR reviewers on #MLsky
November 25, 2024 at 4:47 AM
adam-lg.bsky.social
#statsky #mlsky #blueskai
adam-lg.bsky.social Adam L @adam-lg.bsky.social · May 29
Neat looking paper: "Are Statistical Methods Obsolete in the Era of Deep Learning?" The answer seems to be "no," at least in the case where there's a mathematical structure to the model.

arxiv.org/abs/2505.21723
In the era of AI, neural networks have become increasingly popular for modeling, inference, and prediction, largely due to their potential for universal approximation. With the proliferation of such deep learning models, a question arises: are leaner statistical methods still relevant? To shed insight on this question, we employ the mechanistic nonlinear ordinary differential equation (ODE) inverse problem as a testbed, using physics-informed neural network (PINN) as a representative of the deep learning paradigm and manifold-constrained Gaussian process inference (MAGI) as a representative of statistically principled methods. Through case studies involving the SEIR model from epidemiology and the Lorenz model from chaotic dynamics, we demonstrate that statistical methods are far from obsolete, especially when working with sparse and noisy observations. On tasks such as parameter inference and trajectory reconstruction, statistically principled methods consistently achieve lower bias and variance, while using far fewer parameters and requiring less hyperparameter tuning. Statistical methods can also decisively outperform deep learning models on out-of-sample future prediction, where the absence of relevant data often leads overparameterized models astray. Additionally, we find that statistically principled approaches are more robust to accumulation of numerical imprecision and can represent the underlying system more faithful to the true governing ODEs.
May 29, 2025 at 2:05 AM
shahabbakht.bsky.social
This paper is making the rounds: arxiv.org/abs/2506.21734

A tiny (27M) brain-inspired model trained just on 1000 samples outperforming o3-mini-high on reasoning tasks.

#MLSky 🧠🤖
August 3, 2025 at 2:01 AM
tyrellturing.bsky.social
Join us at this #COSYNE2025 workshop to discuss foundation models for neuroscience!!!

#neuroscience #NeuroAI #MLSky 🧪
averyryoo.bsky.social Avery HW Ryoo @averyryoo.bsky.social · Mar 10
How can large-scale models + datasets revolutionize neuroscience 🧠🤖🌐? We are excited to announce our workshop: “Building a foundation model for the brain: datasets, theory, and models” at @cosynemeeting.bsky.social #COSYNE2025. Join us in Mont-Tremblant, Canada from March 31 – April 1!
March 10, 2025 at 7:56 PM
nicologiso.com
Current situation #datasky #dataBS #MLSky
October 30, 2024 at 4:39 PM
smcgrath.phd
Tagging 🧪 🧬🖥️ #MLSky
gcorso.bsky.social Gabriele Corso @gcorso.bsky.social · Nov 17
Thrilled to announce Boltz-1, the first open-source and commercially available model to achieve AlphaFold3-level accuracy on biomolecular structure prediction! An exciting collaboration with Jeremy, Saro, and an amazing team at MIT and Genesis Therapeutics. A thread!
November 18, 2024 at 1:01 PM
smcgrath.phd
Tagging for 🧪 🩺🖥️ #MLSky 🧠
erictopol.bsky.social Eric Topol @erictopol.bsky.social · Nov 27
When large language models surpass neuroscientists for prediction of results of experiments
nature.com/articles/s41...

"which presages a future where LLMs assist humans
in making discoveries. Our approach is not neuroscience specific and is transferable to other knowledge-intensive endeavours"
November 27, 2024 at 5:16 PM
smcgrath.phd
Tagging for 🧪 #MLSky
December 3, 2024 at 2:48 PM
tedunderwood.com
Flagging this thread on interpretability for #MLSky
colah.bsky.social Chris Olah @colah.bsky.social · Mar 27
How does Haiku generate poetry? It turns out that, after the first line, it generates candidates for the rhyming end of the next line and "writes towards them"!
At the end of the first line of a poem ("He saw a carrot and had to grab it"), Haiku plans candidate words for the end of the next line, and then writes bakcwards.
March 27, 2025 at 8:03 PM
smcgrath.phd
Tagging 🧪 #MLSky
erictopol.bsky.social Eric Topol @erictopol.bsky.social · Jan 1
The 2024 Good Tech Awards, spotlighted by @kevinroose.com for #AI/Science
nytimes.com/2024/12/30/t...
congrats!
@arcinstitute.org @brianhie.bsky.social @patrickhsuu.bsky.social @jameszou.bsky.social and the others for well deserved recognition
January 1, 2025 at 7:38 PM
smcgrath.phd
Tagging 🧪 🩺🖥️ #MLSky
nikitamakarov.bsky.social Nikita Makarov @nikitamakarov.bsky.social · Nov 20
Introducing DT-GPT: showing that LLMs can forecast patient trajectories (1/8)

Pre-print here 👉 medrxiv.org/content/10.1...

Big thanks to Maria Bordukova, @raulrod.bsky.social, Fabian Schmich, Michael Menden, UniMelb, HelmholtzMunich, Roche
November 22, 2024 at 1:19 PM
wimlds.bsky.social
Hello, Bluesky! #MLsky
November 9, 2024 at 6:43 PM
smcgrath.phd
Tagging for 🧪 #MedSky #MLSky #AcademicSky #NurseSky
April 25, 2025 at 6:05 PM
smcgrath.phd
Tagging for 🧪 🩺🖥️ #MLSky
irenetrampoline.bsky.social Irene Chen @irenetrampoline.bsky.social · Dec 17
It's giving Best Paper at the ML for Health Symposium (co-located w NeurIPS)!! 🥳 Congrats to co-authors Emily, Jin, and many others 👏. Check out our work using LLMs to understand liver transplants, esp understudied social and economic factors 🏥💰🏠! #ml4h2024

arxiv.org/pdf/2412.07924
December 17, 2024 at 2:08 AM