James Thornton @JamesTThorn
Research Scientist @Apple ML Research, Paris | Stat / ML PhD Oxford @oxcsml Working on diffusions, optimal transport and sampling jtt94.github.io Paris Joined August 2019-
Tweets260
-
Followers806
-
Following379
-
Likes2K
@neilturkewitz I think the paper has received some negative responses from artists, fueled by provocative Twitter comments by people who haven't read the paper. I want to underline that this is a fundamental *research* paper that solves an interesting mathematical problem and not a product.
Really enjoyed writing this piece with @torfjelde and @vdutor 🙌 Thanks @msalbergo @ValentinDeBort1 @JamesTThorn for your insightful feedback 👌
Really enjoyed writing this piece with @torfjelde and @vdutor 🙌 Thanks @msalbergo @ValentinDeBort1 @JamesTThorn for your insightful feedback 👌
We sadly found out our CTM paper (ICLR24) was plagiarized by TCD! It's unbelievable😢—they not only stole our idea of trajectory consistency but also comitted "verbatim plagiarism," literally copying our proofs word for word! Please help me spread this.
There appears to be a mismatch between publishing criteria in AI conferences and "what actually works". It is easy to publish new mathematical constructs (e.g. new models, new layers, new modules, new losses), but as Apple's MM1 paper concludes: 1. Encoder Lesson: Image…
Exciting News from Open-Sora! 🚀 They've just made the ENTIRE suite of their video-generation model open source! Dive into the world of cutting-edge AI with access to model weights, comprehensive training source code, and detailed architecture insights. Start building your dream…
A variant of mini batch OT flow matching with unbalanced couplings Also very nice to see this in jax: github.com/ExplainableML/… Using: - ott-jax for transport solvers github.com/ott-jax/ott And @PatrickKidger ‘s - equinox for networks - diffrax for ode solvers
A variant of mini batch OT flow matching with unbalanced couplings Also very nice to see this in jax: github.com/ExplainableML/… Using: - ott-jax for transport solvers github.com/ott-jax/ott And @PatrickKidger ‘s - equinox for networks - diffrax for ode solvers
Contrastive learning usually maximize the similarity of exactly 2 views. What happens for >2 views? 🤔 In our ICLR 2024 paper Poly-View Contrastive Learning (PVC) arxiv.org/abs/2403.05490, we find contrasting >2 views outperforms 2-view methods for the same compute 🧵
I am glad this is being recognised. I first saw this advocated by @timudk x.com/timudk/status/… nice to see it written clearly and explained It's also worth noting that sampling from the probability flow ODE of this diffusion model recovers the vector field from flow matching
I am glad this is being recognised. I first saw this advocated by @timudk x.com/timudk/status/… nice to see it written clearly and explained It's also worth noting that sampling from the probability flow ODE of this diffusion model recovers the vector field from flow matching
Please apply if you are a UK undergraduate from under-represented backgrounds interested in exploring what a career in AI research is like. Deadline Feb 17! @GoogleDeepMind has kindly supported the AI internship projects @uniqplusoxford, thank you!
Please apply if you are a UK undergraduate from under-represented backgrounds interested in exploring what a career in AI research is like. Deadline Feb 17! @GoogleDeepMind has kindly supported the AI internship projects @uniqplusoxford, thank you!
I am recruiting a postdoc to work on the foundations of diffusion models. If you are passionate about theory and diffusion models and looking for a unique opportunity to work in an exciting research environment @OxfordStats head over to jobs.ac.uk/job/DFY193/pos… to apply by March 1
Have you ever done a dense grid search over neural network hyperparameters? Like a *really dense* grid search? It looks like this (!!). Blueish colors correspond to hyperparameters for which training converges, redish colors to hyperparameters for which training diverges.
How can you get the best language model for a special task if you want inference to be fast and have little specialization data? We answer this question in our preprint "Specialized Language Models with Cheap Inference from Limited Domain Data"! arxiv.org/abs/2402.01093
New language model work! In practice, LMs often face a double constraint (i) small inference budget + (ii) little application-specific data: (i) means small specialized models for inference; (ii) means using auxiliary generic data e.g. for pretraining 1/2 arxiv.org/abs/2402.01093
Pleased to release the text-to-speech work I developed while at @StabilityAI. 💬 TL;DR - Natural language control of high-fidelity TTS. It’s simple, generalizable, and it sounds better than Audiobox :) text-description-to-speech.com arxiv.org/abs/2402.01912 🧵
Awesome new example - speech and audio generation in MLX. A port of Bark form @suno_ai_ by @chisanchen ! Generate some high quality audio on your laptop. Code: github.com/j-csc/mlx_bark Example:
ICML 2024 call for workshops are open icml.cc/Conferences/20… @BeccaRoelofs, @natschluter, and @andrewgwils are co-chairing. Submit by February 15, 2024, AOE. Help us spread the word! #icml2024
Fast, lightweight data loaders - and DL framework agnostic (also runs on Linux) so a great option for jax if you do not want to have torch / tf requirements
Fast, lightweight data loaders - and DL framework agnostic (also runs on Linux) so a great option for jax if you do not want to have torch / tf requirements
Excited to share AIM 🎯 - a set of large-scale vision models pre-trained solely using an autoregressive objective. We share the code & checkpoints of models up to 7B params, pre-trained for 1.2T patches (5B images) achieving 84% on ImageNet with a frozen trunk. (1/n) 🧵
Apple presents AIM Scalable Pre-training of Large Autoregressive Image Models paper page: huggingface.co/papers/2401.08… paper introduces AIM, a collection of vision models pre-trained with an autoregressive objective. These models are inspired by their textual counterparts, i.e.,…
Michael Hutchinson (@.. @MHutchinson141
687 Followers 335 Following PhD student @OxfordStats / @OXCSML supervised by @yeewhye and @wellingmax. Probabilistic ML, geometric ML and their interestion. Interned @DeepMind @QualcommGabriel Peyré @gabrielpeyre
92K Followers 449 Following @CNRS researcher at @ENS_ULM. One tweet a day on computational mathematics.Sam Power @sp_monte_carlo
17K Followers 7K Following Lecturer in Maths & Stats at Bristol. Interested in probabilistic + numerical computation, statistical modelling + inference. (he / him)Pierre Alquier @PierreAlquier
8K Followers 5K Following Professor of Statistics @ESSEC_AP 🇸🇬 // Previously @RIKEN_AIP 🇯🇵 @ENSAEparis 🇫🇷 @ucddublin 🇮🇪 🇪🇺 // random posts about research & birds photos // 🌈Dan Roy @roydanroy
45K Followers 2K Following ML / AI researcher, emphasis on theory. Research Director and Canada CIFAR AI Chair, @VectorInst Professor, @UofT (Statistics/CS)Kevin Patrick Murphy @sirbayes
42K Followers 334 Following Research Scientist at Google Brain / Deepmind. Interested in Bayesian Machine Learning.Michael Bronstein @mmbronstein
43K Followers 4K Following #DeepMind Professor of #AI @UniofOxford / Fellow @ExeterCollegeOx / ML Lead @ProjectCETI / https://t.co/kZpGpDzYeVyobibyte @y0b1byte
15K Followers 2K Following Kurin ViTaly, senior research scientist @IsomorphicLabs, ML PhD from @UniofOxford on RL, Multitask learning & GraphsJean-François Ton @jeanfrancois287
909 Followers 782 Following Senior Research Scientist @BytedanceTalk working on Responsible AI prev. @oxcsml @UnioxOxford, @amazon, @apple, @bloomberg All opinions are my ownArash Vahdat (hiring) @ArashVahdat
8K Followers 806 Following Principal scientist and research manager @nvidia research, leading forward-looking fundamental generative AI research efforts, views are my own.Sander Dieleman @sedielem
50K Followers 2K Following Research Scientist at Google DeepMind. I tweet about deep learning (research + software), music, generative models (personal account).Chris J. Maddison @cjmaddison
18K Followers 2K Following Asst. Prof. in Machine Learning at UofT and #LongCOVID patient.Andreas Kirsch 🇮�.. @BlackHC
9K Followers 5K Following Past: 🧑🎓 DPhil @AIMS_oxford @ExeterCollegeOx @UniofOxford (4.5yr) 🧙♂️ RE @DeepMind (1yr) 📺 SWE @Google (3yrs) 🎓 @TU_Muenchen 👤 Fellow @nwspkAdam Foster @AdamEFoster
551 Followers 189 Following Senior Researcher at Microsoft Research AI4Science. Previously Oxford PhD in machine learningGuan-Horng Liu @guanhorng_liu
696 Followers 321 Following ML PhD @GeorgiaTech 🚀 • previously MS @CMU_robotics • Schrödinger Bridge / diffusion / stochastic optimal control • intern @MetaAI @NVIDIAAIMiguel Angel Bautista @itsbautistam
2K Followers 181 Following I am a research scientist (currently @ Apple ML Research) seeking a grand unification of generative modeling 🇪🇸🇺🇸Charlotte Bunne @_bunnech
3K Followers 484 Following PostDoc at @Genentech and @Stanford and Incoming Assistant Professor at @EPFL in Computer Science and Life Sciences.Julius Berner @julberner
529 Followers 265 Following Postdoc @caltech | PhD @univienna | former research intern @MetaAI and @nvidia | bridging theory and practice in deep learningMathieu Alain @miniapeur
19K Followers 2K Following Researching @ai_ucl. Co-organises @uclcsml and @logconference. FR, EN, trying ES. 🇹🇼🇨🇦🇬🇳🇺🇸🇩🇴🇫🇷🇪🇸🇬🇧🇿🇦Alejandra Avalos @AleAviP
2K Followers 1K Following 🇲🇽|Univ.Ass. @tu_wien🇦🇹|Member @HarvardMIT_CRS🇺🇸|Chair @j_ISBA📈|Ex @dfcidatascience🇺🇸@DataScienceFLR🇮🇹|Alumna @warwickstats @OxfordStats 🇬🇧Sandra Joelson @sandr_joels
35 Followers 5K FollowingHannahPaul @Y0TQ8l9o5f8W4y
0 Followers 89 FollowingLucile Maye @lucile_may41115
82 Followers 5K FollowingMichael Kirchhof @mkirchhof_
652 Followers 287 Following PhD student @uni_tue. Working on large-scale #uncertainty quantification in #machinelearningArif Ahmad @arif_ahmad_py
278 Followers 7K Following All things AI, Computer Science and Circuits! Prev. @GoogleAIAngelos Katharopoulos @angeloskath
2K Followers 236 Following Machine Learning Research @Apple. Previously PhD student at @idiap_ch and @EPFL. Interested in all things machine learnableAbbas Mammadov @AbbasMammadov11
6 Followers 35 FollowingRishabh Anand 🧬 @rishabh16_
5K Followers 1K Following multiplying matrices @NUSingapore • geometric DL + generative modelling for proteins, RNA, and drug discovery @Cambridge_CL 🛠John Wong @ChiHoWONG19
43 Followers 141 FollowingPaulina Szymczak @szymczak_pau
376 Followers 2K Following Computational biology by day, SciArt by night. PhD student at @ewa_szczurek's lab @HelmholtzMunich. Single-cell multiomics & antimicrobial peptides (she/her).Sebastian @Sebastian_ae_
22 Followers 230 FollowingMichael E. Sander @m_e_sander
721 Followers 199 Following Ph.D. student @ENS_ulm, with @gabrielpeyre and @mblondel_ml.Laiba Rester @LaibaRest
48 Followers 5K FollowingKostas Tsampourakis @KTsampourakis
78 Followers 650 Following Stats PhD @EdinburghUni. Work on Prob ML, focus on SSM, Bayesian state estimation/filtering & learning dynamics. https://t.co/2xjHceuRKqMartin Fan @perfectoid_ai
395 Followers 8K FollowingLyla-rose Depadua @DepaduRos
73 Followers 5K FollowingPaloma Yengich @yengich54276
93 Followers 5K FollowingMason Minot @MasonMinot
60 Followers 382 Following PhD Student ETH Zurich. Interested in ML and drug development. Previously @Genentech, @CornellBaran Hashemi @Rythian47
482 Followers 4K Following Postdoc at ORIGINS Cluster | @TU_Muenchen , Ex @LMU_Muenchen, #AI4Science, #AI4MathNaomi Moak @NaoMoak
48 Followers 5K FollowingMartin Jørgensen @JorgensenMart
483 Followers 602 Following Postdoctoral researcher, University of Helsinki, Statistical Machine LearningEva Louise Marie Gabr.. @e681554349
9 Followers 3K Followinggylns @glovepm
3 Followers 437 Followingzihan charlie @zihan294
12 Followers 77 FollowingJannis Bolik @BolikJannis
10 Followers 92 FollowingK @spearmintfresh1
18 Followers 149 FollowingMarco Matthies @MarcoMatthies
92 Followers 2K Following Interested in math, programming, computational biology, AI, and investing.Su @BilgeSuuuu
1K Followers 1K FollowingRui-Yang Zhang @ruiyang3927
25 Followers 96 FollowingJoy Chopra @joy_chopra2
20 Followers 1K FollowingMahdi Mehmanchi @_Mahdi_M_M97
32 Followers 354 Following AI Graduate Student @ University of Tehran Interested in deep learning theory, optimization, and generative modelsMatthew Johnson @SingularMattrix
12K Followers 3K Following Researcher at Google Brain. I work on JAX (https://t.co/UGa5tGfinF).daiki-ko @daikiko_0422
248 Followers 1K Following Assistant Professor (Ph.D. in Engineering) : Cheminformatics / Machine Learning for MoleculesFelipe Cisternas A. @ftcister
15 Followers 469 Following MSc. in Computer Science at Universidad Tecnica Federico Santa Maria. Focused on Data Science, Machine Learning and Artificial Intelligence.Putra Manggala @pmangg
503 Followers 4K Following researcher @amlabuva, previously @shopify, @guavus, @adgear, @mcgillu. Not fun at parties.Jiatao Gu @thoma_gu
3K Followers 2K Following Machine Learning Researcher at @Apple ML Research (MLR) based in NYC | ex-FAIRer | PhD from HKU | Research on Generative AI for multimodalities. また日本語もできます。Shyamgopal Karthik @ShyamgopalKart1
225 Followers 719 Following PhD candidate @uni_tue @ml4science , former intern @naverlabseurope , Master's from @iiit_hyderabadMichael Hutchinson (@.. @MHutchinson141
687 Followers 335 Following PhD student @OxfordStats / @OXCSML supervised by @yeewhye and @wellingmax. Probabilistic ML, geometric ML and their interestion. Interned @DeepMind @QualcommGabriel Peyré @gabrielpeyre
92K Followers 449 Following @CNRS researcher at @ENS_ULM. One tweet a day on computational mathematics.Sam Power @sp_monte_carlo
17K Followers 7K Following Lecturer in Maths & Stats at Bristol. Interested in probabilistic + numerical computation, statistical modelling + inference. (he / him)AK @_akhaliq
310K Followers 3K Following AI research paper tweets, ML @Gradio (acq. by @HuggingFace 🤗) dm for promo follow on Hugging Face: https://t.co/q2Qoey80GxPierre Alquier @PierreAlquier
8K Followers 5K Following Professor of Statistics @ESSEC_AP 🇸🇬 // Previously @RIKEN_AIP 🇯🇵 @ENSAEparis 🇫🇷 @ucddublin 🇮🇪 🇪🇺 // random posts about research & birds photos // 🌈Dan Roy @roydanroy
45K Followers 2K Following ML / AI researcher, emphasis on theory. Research Director and Canada CIFAR AI Chair, @VectorInst Professor, @UofT (Statistics/CS)Kevin Patrick Murphy @sirbayes
42K Followers 334 Following Research Scientist at Google Brain / Deepmind. Interested in Bayesian Machine Learning.Michael Bronstein @mmbronstein
43K Followers 4K Following #DeepMind Professor of #AI @UniofOxford / Fellow @ExeterCollegeOx / ML Lead @ProjectCETI / https://t.co/kZpGpDzYeVyobibyte @y0b1byte
15K Followers 2K Following Kurin ViTaly, senior research scientist @IsomorphicLabs, ML PhD from @UniofOxford on RL, Multitask learning & GraphsJean-François Ton @jeanfrancois287
909 Followers 782 Following Senior Research Scientist @BytedanceTalk working on Responsible AI prev. @oxcsml @UnioxOxford, @amazon, @apple, @bloomberg All opinions are my ownFrançois Chollet @fchollet
470K Followers 769 Following Deep learning @google. Creator of Keras. Author of 'Deep Learning with Python'. Opinions are my own.Arash Vahdat (hiring) @ArashVahdat
8K Followers 806 Following Principal scientist and research manager @nvidia research, leading forward-looking fundamental generative AI research efforts, views are my own.Sander Dieleman @sedielem
50K Followers 2K Following Research Scientist at Google DeepMind. I tweet about deep learning (research + software), music, generative models (personal account).Ben Poole @poolio
17K Followers 1K Following research scientist at google brain. phd in neural nonsense from stanford.Stat.ML Papers @StatMLPapers
20K Followers 0 Following Unofficial updates of statistical machine learning papers on arXivAdam Foster @AdamEFoster
551 Followers 189 Following Senior Researcher at Microsoft Research AI4Science. Previously Oxford PhD in machine learningJascha Sohl-Dickstein @jaschasd
19K Followers 625 Following Member of the technical staff @ Anthropic. Most (in)famous for inventing diffusion models. AI + physics + neuroscience + dynamics.Durk Kingma @dpkingma
35K Followers 348 Following Deep learning, mostly generative models. Prev. Google Brain/DeepMind, founding team @OpenAI. Inventor of the VAE, Adam optimizer, among other things. ML PhD.Patrick Kidger @PatrickKidger
9K Followers 192 Following 🧪BioML @ https://t.co/04dWAWzCyl 🧑💻Prev. Google X, Oxford. 📚Neural ODE textbook: https://t.co/ODOKWjub5k 🤖Open JAX ecosystem: https://t.co/8kXzaG9XVfAhmad Al-Dahle @Ahmad_Al_Dahle
4K Followers 53 Following #Girldad of twins. Leading GenAI @ Meta (llama, imagine, meta ai and more)Jiaming Song @baaadas
5K Followers 992 Following Chief Scientist @LumaLabsAI. Working on visual generative AI. Were @NVIDIA @Stanford @OpenAI @MetaAIMichael E. Sander @m_e_sander
721 Followers 199 Following Ph.D. student @ENS_ulm, with @gabrielpeyre and @mblondel_ml.Laurent Sifre @laurentsifre
1K Followers 411 Following Research Scientist @DeepMind since 2014. Worked on #AlphaGo #AlphaFold and #AlphaStar, now focused on #NLP at scale.Felix Petersen @FHKPetersen
1K Followers 195 Following Machine learning researcher, postdoc, 24 y/o. Investigating differentiable algorithms etc. @Stanford @StanfordAILab.Nando de Freitas 🏳.. @NandoDF
97K Followers 659 Following I research intelligence to understand it and to harness it wisely. Part of AlphaGo tuning, AlphaCode, learning to learn, Lyria, Imagen2, Gato, rGemmaJesse Engel @jesseengel
9K Followers 59 Following Guitarist, Researcher Google DeepMind. Opinions are my own. @[email protected]Mathieu Blondel @mblondel_ml
9K Followers 421 Following Research scientist at Google DeepMind. Current research interests: differentiable programming, LLMs, Transformers.Julius Berner @julberner
529 Followers 265 Following Postdoc @caltech | PhD @univienna | former research intern @MetaAI and @nvidia | bridging theory and practice in deep learningGael Varoquaux @GaelV.. @GaelVaroquaux
22K Followers 318 Following Research & code: Research director @inria ►Data, Health, & Computer science ►Python coder, (co)founder of @scikit_learn & joblib ►Art on @artgael ►Physics PhDDominik Klein @Dominik1Klein
257 Followers 154 Following @EllisforEurope PhD student, Student Researcher @Apple. Previously MSc @OxfordStats. Interested in Machine Learning, Single-Cell Genomics, and People.Matthew Johnson @SingularMattrix
12K Followers 3K Following Researcher at Google Brain. I work on JAX (https://t.co/UGa5tGfinF).Théo Uscidda @theo_uscidda
21 Followers 98 Following PhD @ENSAEparis, working on generative modeling & optimal transport with @CuturiMarco.Xiaohua Zhai @XiaohuaZhai
3K Followers 208 Following Senior Staff Researcher @GoogleDeepMind team in ZürichShyamgopal Karthik @ShyamgopalKart1
225 Followers 719 Following PhD candidate @uni_tue @ml4science , former intern @naverlabseurope , Master's from @iiit_hyderabadNathan Benaich @nathanbenaich
51K Followers 32K Following solo member of investment staff @airstreet, brewing ambition @airstreetcafe, next token predictor @airstreetpressRob Brekelmans @brekelmaniac
382 Followers 220 Following postdoc @vectorinst (phd @usc_isi, intern @googledeepmind) what’s next?Joseph Watson @_JosephWatson
3K Followers 108 Following Co-Founder @ Xaira Therapeutics Former Baker Lab PostDoc at @UWproteindesign. Interested in generative modelling for molecular design. Views my own.Hila Manor @hila8manor
66 Followers 70 Following PhD student @TechnionLive | Interested in uncovering hidden knowledge in ML models | ML + Music + Uncertainty = ♥Saining Xie @sainingxie
14K Followers 1K Following researcher in #deeplearning #computervision | assistant professor at @NYU_Courant @nyuniversity | previous: research scientist @metaai (FAIR) @UCSanDiegoVimal Thilak🦉🐒 @AggieInCA
372 Followers 597 Following Proverbs 17:28. I’m not learned. A deep delver.Shreyas Padhy @shreyaspadhy
260 Followers 553 Following PhD student at the University of Cambridge. Ex @GoogleAI Resident, @jhubme and @iitdelhi. I like the math of machine learning & neuroscience. Also DnD.Angelos Katharopoulos @angeloskath
2K Followers 236 Following Machine Learning Research @Apple. Previously PhD student at @idiap_ch and @EPFL. Interested in all things machine learnableArthur Mensch @arthurmensch
40K Followers 873 Following Co-founder and CEO @MistralAI. Apply https://t.co/yHGRZAtjcxMarin Vlastelica 🤖.. @vlastelicap
1K Followers 1K Following Final year PhD @ Max Planck Institute for Intelligent Systems 🤖 | All things ML 🎲 | ex. @DeepMind, @amazon 🇭🇷🇩🇪🎸🏀🎾Rivers Have Wings @RiversHaveWings
31K Followers 224 Following AI/generative artist. Writes her own code. Absolute power is a door into dreaming.Piotr Dabkowski @dabkowski_piotr
855 Followers 82 Following Co-Founder & Research at ElevenLabs (@elevenlabsio) ⑪Michal Valko @misovalko
5K Followers 2K Following Llama @AIatMeta Paris & Inria & MVA - Ex: Gemini and BYOL @GoogleDeepMindSara Sangalli @salusanga_
56 Followers 121 Following PhD candidate at Computer Vision Lab, ETH ZürichSasha Rush @srush_nlp
52K Followers 464 Following Professor, Programmer in NYC. Cornell Tech, Hugging Face 🤗 https://t.co/cZl0wTfqGzJonathan Crabbé @JonathanICrabbe
205 Followers 226 Following 🎓PhD in ML Interpretability @Cambridge_Uni ⚙️ Ex-Research Scientist Intern @Apple & @MSFTResearch 💡 Interested in Interpretability, Robust ML & GenAISam Bond-Taylor @sambondtaylor
157 Followers 168 Following Researcher in Health Futures at @MSFTResearch. Previously PhD in deep generative models @comp_sci_durham. He/him.koray kavukcuoglu @koraykv
8K Followers 84 Following VP of Research and Technology at Google DeepMindStephan Mandt @StephanMandt
2K Followers 556 Following ML Professor @UCIrvine, previously @blei_lab, @Princeton. #GenerativeAI, #Compression, #AI4Science. Program Chair @aistats_conf 2024; General Chair AISTATS 2025Ruiqi Gao @RuiqiGao
5K Followers 511 Following Research scientist @Google DeepMind. Generative modeling, representation learning.Liliane Momeni @LiliMomeni
744 Followers 714 Following PhD @Oxford_VGG w/ Prof Zisserman (grad. early ’24) • @Google PhD Fellowship in Perception • prev. Research Intern @Google & @MetaAlaa El-Nouby @alaaelnouby
514 Followers 656 Following PhD Student at @MetaAI and @Inria. Studied my MSc at @VectorInst and @UofG. Previously interned at @MSFTResearch and @Apple. Egyptian 🇪🇬Alaa El-Nouby @alaa_nouby
521 Followers 302 Following Research Scientist at @Apple. Previous: @Meta (FAIR), @Inria, @MSFTResearch, @VectorInst and @UofG . Egyptian 🇪🇬 Deprecated twitter account: @alaaelnoubyChenlin Meng @chenlin_meng
8K Followers 833 Following Co-founder & CTO @pika_labs | ex @StanfordAILab @StanfordIn case you are wondering, this paper proves that, in general, diffusion models do not define optimal transport maps. The proof is not straightforward though (diffusion maps are optimal maps in 1D, for radial measure and for Gaussians ...) cvgmt.sns.it/media/doc/pape…
The optimal computation of gradients for the composition of functions is an optimal parenthesis problem. Forward and backward (backpropagation) are two extreme cases. Backward is optimal for scalar-valued functions. link.springer.com/article/10.100… en.wikipedia.org/wiki/Matrix_ch…
I have to say it because @awnihannun is quick to give credit to others but doesn’t take much for himself. This performance improvement largely comes from his relentless hunting down of every kind of overhead in MLX the past weeks. Kudos!!!
MLX 0.10 → 0.11, faster generation across model sizes and machines. tokens-per-second for 4-bit models:
📢📢 Align Your Steps: Optimizing Sampling Schedules in Diffusion Models research.nvidia.com/labs/toronto-a… TL;DR: We introduce a method for obtaining improved sampling schedules for diffusion models, resulting in better samples at the same computation cost. (1/5)
🤔 So everyone is talking about SSMs, but little theory has been developed to describe their learning abilities. ⚡️In our recent paper arxiv.org/abs/2402.19047, led by @MucaCirone, we give theoretical grounding to the field of SSMs using tools from Rough Path Theory! 🧵(1/6)
@neilturkewitz I think the paper has received some negative responses from artists, fueled by provocative Twitter comments by people who haven't read the paper. I want to underline that this is a fundamental *research* paper that solves an interesting mathematical problem and not a product.
Did you know you can train a good generative model, even if your training data is corrupted or noisy? Our paper, recently accepted to @TmlrOrg, does exactly that. 🧵 openreview.net/forum?id=BRl7f…
Consistent Diffusion Meets Tweedie. Our latest paper introduces an exact framework to train/finetune diffusion models like Stable Diffusion XL solely with noisy data. A year's worth of work breakthrough in reducing memorization and its implications on copyright 🧵
@dereklim_lzh #NVIDIA #ICLR2024 spotlight paper: Graph Metanetworks We give a framework for processing neural nets with other neural nets, improving expressiveness and performance. My favorite use is generating the weights of implicit neural representations -- ex., in 3D generation
@BigAmeya excalidraw.com It's a super simple, neat tool! @james_r_lucas did most of the figures
🚀 Excited to introduce my internship work at @Apple MLR : Many-to-many Image Generation with Auto-regressive Diffusion Models (arxiv.org/abs/2404.03109). Exploring the paradigm for domain-general multi-image to multi-image generation.
Really enjoyed writing this piece with @torfjelde and @vdutor 🙌 Thanks @msalbergo @ValentinDeBort1 @JamesTThorn for your insightful feedback 👌
Along with @MathieuEmile and @vdutor we've cooked up a gentle introduction to flow matching, a recent method for efficiently training continuous normalizing flows (CNFs)! Hope you find it interesting! mlg.eng.cam.ac.uk/blog/2024/01/2… 1/2
And big thanks @msalbergo @ValentinDeBort1 @JamesTThorn for your insightful and valuable feedback!
The indomitable Miriam Margolyes OBE has a message in support of the Jewish Council! She calls for all of us Jews to “shout, beg, scream for a ceasefire”
I'd estimate that I spend at least 4-5x the paid hours on students that I supervise. we all know this but: it's terrible for students. it's terrible for junior academics. and it should be terribly embarrassing for the university.
proposal: @NeurIPSConf @icmlconf @iclr_conf @UncertaintyInAI @aistats_conf @ECMLPKDD (and more) why don't you sync to cap the max number of papers that any author can submit in a year to all of you combined. Maybe set it to 10. this will help with the #review crisis in #ML #AI
Super excited to share Universal-1, our new SOTA multilingual automatic speech recognition model - trained in Jax 🚀!
Introducing Universal-1, our most powerful speech recognition model to date. Trained on over 12.5 million hours of multilingual audio data, Universal-1 achieves best-in-class speech-to-text accuracy across English, Spanish, French, and German. assemblyai.com/research/unive…
10 years ago to the day, I published my first ML-related blog post: sander.ai/posts/ My blogging has been very sporadic over the years, but sharing what I've learnt has been very rewarding, and probably a pretty good career move as well😁 I highly recommend it!
@tokenpilled65B I do hope to learn Pallas at some point. Maybe we can even hook their triton backend into the visualizer. Might put together some Jax puzzles over the summer with some PyTree nonsense.