logprob @logprob
phd student, deep learning Joined January 2019-
Tweets220
-
Followers130
-
Following3K
-
Likes1K
New blog post: Implicit ODE Solvers Are Not Universally More Robust than Explicit ODE Solvers, Or Why No ODE Solver is Best Talks about how "robust" methods can be less robust in practice. Justifies hundreds of methods in #julialang #sciml stochasticlifestyle.com/implicit-ode-s…
Does a smaller latent space lead to worse generation in latent diffusion models? Not necessarily! We show that LDMs are extremely robust to a wide range of compression rates (10-1000x) in the context of physics emulation. We got lost in latent space. Join us 👇
Really cool to see how the bigger-doesn't-mean-better trends in language models also seem to hold up for science models Hope this means that computational physics will no longer be solely in the domain of enormous HPC clusters
Really cool to see how the bigger-doesn't-mean-better trends in language models also seem to hold up for science models Hope this means that computational physics will no longer be solely in the domain of enormous HPC clusters
there is the urgent need for understanding the internal behavior of deep nets. (I am at a numerical analysis conference, tired of listening to talks blindly applying a random architecture to Navier Stokes with periodic boundary conditions)
The big breakthrough for convnets was the first GPU-accelerated CUDA implementation, which immediately started winning first place in image classification competitions. Remember when that happened? I do. That was Dan Ciresan in 2011
The big breakthrough for convnets was the first GPU-accelerated CUDA implementation, which immediately started winning first place in image classification competitions. Remember when that happened? I do. That was Dan Ciresan in 2011
'Mathematics compares the most diverse phenomena and discovers the secret analogies that unite them.' -- Joseph Fourier
Every organism yearns to maximize its mutual information with the future. One way is to live forever, but the optimal way is to be fruitful (genetically and memetically) and multiply your bits (ideas and genes) so they take up more of the future lightcone.
don't know how this LLM thing will be, I'd rather bet on SciML
1 year later, still far away
i'm increasingly convinced that "transformative ai" is going to look like an abundance of specialized models for everything from drug design to weather sims to robotics to supply chains, not one agent to rule them all. we're going to need a lot more ai researchers
humans are imitation machines, the base level of “intelligence” has already been achieved
humans are imitation machines, the base level of “intelligence” has already been achieved
it got navier-stokes
@cboyack @DiscoverPraxis The point of college is not to prepare you for "professional success." It's to improve your mind. Nor are these distinct paths. The people who achieve the greatest "professional success" are those who think of college as more than job training.
Upgraded from Llama 3 to Qwen3 as my go-to model for research experiments, so I implemented qwen3 from scratch: github.com/rasbt/LLMs-fro… Trade-off: Qwen3 0.6B is deeper (28x vs 16x layers) & slower than the wider Llama 3 1B but more memory efficient due to fewer params
New paper on the generalization of Flow Matching arxiv.org/abs/2506.03719 🤯 Why does flow matching generalize? Did you know that the flow matching target you're trying to learn **can only generate training points**? with @Qu3ntinB, Anne Gagneux & Rémi Emonet 👇👇👇
Is exponentially accumulating errors what LeCun said all along? That part might be true, but it doesn't mean LLMs are 'doomed'. Dropping the error rate from 10% to 1% (per 10min) makes 10h tasks possible. In practice, the error rate has been halving every 4 months(!). In…
Is exponentially accumulating errors what LeCun said all along? That part might be true, but it doesn't mean LLMs are 'doomed'. Dropping the error rate from 10% to 1% (per 10min) makes 10h tasks possible. In practice, the error rate has been halving every 4 months(!). In… https://t.co/Za2FTyzIsB
guessing how much energy does chatgpt burns to spit out those emojis
btw, i wrote a post about "how to scale" based on what i've learned over the past few months. it covers muP, HP scaling laws, and some stuffs. would be happy to get any feedback or discussion. (it's pretty verbose and no TL;DR, sorry lol) howtoscalenn.github.io
Jewel @j_wiese90
309 Followers 3K Following
Brenda @cordano_brenda3
271 Followers 3K Following
MarciaChristian @IyXP2JA1YE8h77
2 Followers 159 Following
James Reed @jamesr66a
894 Followers 3K Following Deep Learning, Systems, Compilers, Signals, Languages (Natural & PL). Multimedia Lead @FireworksAI_HQ prev PyTorch @Meta. Tweets my own, mostly animal pics
Dovie Hudson @DovieHudso34919
86 Followers 4K Following
Paula @0lOWnYBT97uVUBQ
11 Followers 885 Following A woman with a voice is, by definition, a strong woman.
Bianca @Talsor326216
17 Followers 904 Following Stay strong. Make them wonder how you’re still smiling.
Marigold @o05C3Q8LhJQ73
26 Followers 900 Following
Holly @Duiqhou07517
14 Followers 886 Following Sometimes the most productive thing you can do is relax.
Mirie @Mirie5266768
7 Followers 691 Following
Elliott Price @elliott_pr37951
38 Followers 2K Following
Anya Jast @AnyaJast81111
36 Followers 2K Following
Buho @Buho6448
48 Followers 2K Following
Oozroobo @Oozroobo4001
95 Followers 2K Following
rank decomposition @rankdim
772 Followers 310 Following my machine is not learning | discord @ rank.dim | email @ req
Ferui @Ferui72591
72 Followers 2K Following
Jack D. Carson @mtlushan
2K Followers 918 Following eecs&physics @mit - omniscience enthusiast - training big biology models @mit_csail @mskcancercenter
Nimit Kalra @ ICML 20... @qw3rtman
1K Followers 927 Following research @haizelabs, prev @citadel, @utaustin currently feynman technique-ing my way through life
Jacob Zietek @JacobZietek
1K Followers 2K Following ml @personaaiinc | embodied ai | ex amazon, armada ai, shield ai | operating in latent space | 🇺🇸🇵🇱
Rebecca @a_rebecca43
239 Followers 3K Following
Pierre Chambon @PierreChambon6
807 Followers 2K Following NLP/Code Generation PhD at FAIR (Meta AI) and INRIA - previously researcher at Stanford University - MS Stanford 22’ - Centrale Paris P2020
Mohmed Mostafa @M0hmed_Mostafa
7 Followers 336 Following
Pierfrancesco Beneven... @PierBeneventano
520 Followers 817 Following Postdoc at @MIT | ex PhD student @Princeton | Exploring how to train AIs and their interaction with the world, while brewing my espresso.
Maria @molacek_maria
290 Followers 3K Following
Sichu Lu(Sichu.Lu218@... @lu_sichu
3K Followers 6K Following nlab fan account, arxiv surveyor, pubmed enjoyer, two culture bridger, vacuous high gossiper, dearth of any domain expertise, reluctant g theorist, gpu poor,
Amy @amy_richie_
159 Followers 3K Following
Jiahan Li @Ced3LJH
259 Followers 957 Following
Thidares @Thidareseb5tSL
30 Followers 2K Following
Sotha @SothaWbd
10 Followers 621 Following
Clara @clara40keys
327 Followers 3K Following
Will Redman @wredman4
372 Followers 509 Following Incoming assist. prof. JHU ECE | Koopman operators and hippocampi | He/Him | You know what the French c'est - "Hate baguettes hate"
Ashish Kapoor @akapoor_av8r
5K Followers 364 Following Building general purpose robotics intelligence @genrobotics_ai | Aviator
Ting Chen @tingtang222
317 Followers 517 Following PhD student @Columbia Stats. previously @urichmond CS, @EdinburghUni informatics. interested in Bayesian statistics and machine learning
Peter Zaika @PeterZaika
16 Followers 74 Following Phd Student in Mathematics at the University of Cambridge
a @yihyunCS
2K Followers 4K Following schizo-shitposter the child asked me if god exists. i said 'not yet.' prev @nasa
Fearseigh @FearseighE8z
89 Followers 2K Following
Clayton Thorrez @cthorrez
1K Followers 2K Following Rating systems and paired comparison experimentation enjoyer @lmarena_ai
Sid 🖥️🧬 refle... @hopelesslystoic
1K Followers 4K Following Intern @ frontier tower SF || any/all || 25 || Bioinformatician || Holistically Nerdy 🌼
Ari Morcos @arimorcos
7K Followers 2K Following CEO and Co-founder @datologyai working to make it easy for anyone to make the most of their data. Former: RS @AIatMeta (FAIR), RS @DeepMind, PhD @PiN_Harvard.
Jeremy Cohen @deepcohen
4K Followers 928 Following Research fellow at Flatiron Institute, working on understanding optimization in deep learning. Previously: PhD in machine learning at Carnegie Mellon.
Yilun Xu @xuyilun2
2K Followers 348 Following AI @Meta . Prev: World Sim @GoogleDeepMind, @NVIDIA, PhD @MIT_CSAIL, BS @PKU1898 . views are my own
François Rozet @FrancoisRozet
669 Followers 136 Following datamancer, generative models, bayesian inference, dynamical systems, open-source software, phd with @glouppe
Laτenτ Holdings @latentholdings
3K Followers 16 Following Acceleraτing τhe digiτal τransformaτion of inτelligence. // @taoapp_, @learnbittensor, @taohash, @traininghone
Hone Training @traininghone
375 Followers 10 Following High Compression AI Pretraining · Subnet 5 on Biττensor · powered by @manifoldlabs & @latentholdings
Manifold @manifoldlabs
4K Followers 9 Following A Decentralized Frontier AI Lab @TargonCompute & @TrainingHone
τargon @TargonCompute
882 Followers 6 Following High-Speed Decentralized Compute Cloud · Subnet 4 on Biττensor · powered by @manifoldlabs
Edd @erla_ndpg
207 Followers 1K Following Ex Computer Science student from Indonesia Only create X because AI Researchers are on X
James Reed @jamesr66a
894 Followers 3K Following Deep Learning, Systems, Compilers, Signals, Languages (Natural & PL). Multimedia Lead @FireworksAI_HQ prev PyTorch @Meta. Tweets my own, mostly animal pics
Sheng Shen @shengs1123
2K Followers 551 Following @xAI | Prev. 🦙MetaAi; MSFTResearch, allen_ai, GoogleDeepMind; PhD @berkeley_ai
Liangchen Luo @LiangchenLuo
4K Followers 122 Following @xAI reasoning; ex @GoogleDeepMind. B.Sc. @PKU1898. Opinions are my own.
Lei Gong @gonglei89
19K Followers 139 Following Mostly tweet about politics, econ, tech, and China. Find me @ https://t.co/Y5mI4BJnVE
Mercor @mercor_ai
10K Followers 20 Following We use AI to understand human ability and match talent with the opportunities they're best suited for.
Microsoft AI @MicrosoftAI
2K Followers 6 Following We build responsible AI to empower people | https://t.co/A80GqUp4Kb
Analysis Fact @AnalysisFact
133K Followers 19 Following Daily tweets about real and complex analysis and related topics. From @JohnDCook.
Math Cafe @Riazi_Cafe_en
37K Followers 115 Following The goal of this account is to encourage mathematical thinking. Farsi account: @Riazi_Cafe
Francesco Bertolotti @f14bertolotti
802 Followers 130 Following Postdoctoral researcher at the university of Milan
Mechanize @MechanizeWork
6K Followers 1 Following We're a software company building RL environments to power the full automation of the economy.
Mor Geva @megamor2
2K Followers 529 Following Assistant Professor at @TelAvivUni and Research Scientist at @GoogleResearch; previously postdoc at @GoogleDeepMind and @allen_ai
Rachel @tolstoybb
28K Followers 1K Following
Mihaela van der Schaa... @MihaelaVDS
9K Followers 402 Following John Humphrey Plummer Professor of Machine Learning, Artificial Intelligence and Medicine @Cambridge_Uni, Fellow @royalsociety
XiaomiMiMo @XiaomiMiMo
613 Followers 3 Following
BAAI @BAAIBeijing
3K Followers 228 Following Beijing Academy of Artificial Intelligence. AI Science and Technology Innovation for Global Sustainable Development.
Ant Ling @AntLing20041208
112 Followers 84 Following A series of open-sourced large models from Ant Group,Ling for LLM, Ring for Reasoning LLM,Ming for MLLM. See us at inclusionAI.
InclusionAI @InclusionAI666
112 Followers 64 Following Open-source projects conducted by Ant Group，including Ling,AReal,AWorld. Dedicated our efforts towards AGI,guided by fairness, transparency, and collaboration.
Yuxiang Wei @YuxiangWei9
710 Followers 263 Following PhD candidate @IllinoisCDS | Researcher @AIatMeta (Meta FAIR). Code LLM training.
John Jumper @JohnJumperSci
1K Followers 0 Following
Jesús Fernández-Vil... @JesusFerna7026
22K Followers 68 Following Howard Marks Presidential Professor of Economics at the University of Pennsylvania and John H. Makin Visiting Scholar at the AEI. All opinions are my own.
TBPN @tbpn
102K Followers 921 Following Technology's daily show. Hosted by @johncoogan and @jordihays. Streaming live 11AM-2PM PT every weekday and available on Apple, Spotify, and YouTube.
Johannes Hagemann @johannes_hage
8K Followers 2K Following co-founder/cto @PrimeIntellect | decentralized AI, longevity, techno-optimism
ML Collective @ml_collective
14K Followers 42 Following We are an independent nonprofit organization that believes collaboration opportunities and research training should be openly accessible and free.
Bryan Johnson @bryan_johnson
646K Followers 759 Following Conquering death will be humanity’s greatest achievement.
Andrew D. Huberman, P... @hubermanlab
1.6M Followers 2K Following Professor of Neurobiology and Ophthalmology at Stanford Medicine • Host of Huberman Lab • Focused on science and health research and public education
Ion Stoica @istoica05
5K Followers 20 Following Professor at UC Berkeley, co-founder of Databricks, Anyscale, LMArena, Conviva.
yung macro 宏观年�... @apralky
24K Followers 626 Following rates trader learning about markets & society... gen z supremacist & technocratic elitist. not financial advice
Arc Jax @arcjax7
160 Followers 2K Following
Leo Zhang @leoeleoleo1
257 Followers 389 Following PhD student @OxfordStats working on generative models
Divam Gupta @divamgupta
10K Followers 131 Following Building super-tiny AI models that (hopefully) think • YC S25 • Made @DiffusionBee • Previously: AI research @Microsoft, @CarnegieMellon, @Meta
Georgia Channing @cgeorgiaw
679 Followers 208 Following AI4Science @ 🤗, PhD @OxfordTVG — sharing the world of science
Sina @SinaHartung
11K Followers 750 Following slightly less attractive cofounder @AskEureka: we’re replacing all doctors with AI. I tweet abt healthcare and tech, prev @Harvard @Google @BCG, dm to say hi :)Trends for United States
You might like
