Mojtaba Vàlipour @ValipourMojtaba
CS PhD at @UWaterloo, Founding Engineer at Coastal Carbon and part time Researcher at @huawei Noah’s Arc Lab, prev enjoyed my time at @oraclelabs, & @CVC_UAB Waterloo, Ontario Joined February 2013-
Tweets1K
-
Followers404
-
Following3K
-
Likes2K
[Download 601-page PDF eBook] Mathematical Introduction to #DeepLearning — Methods, Implementations, and Theory: arxiv.org/abs/2310.20360 ———— #AI #Mathematics #BigData #DataScience #Algorithms #MachineLearning #Calculus #LinearAlgebra #DataScientists
arxiv.org/abs/2404.15702 Doesn't seem to perform incredibly, but they have a lot of neat details about the training pipeline
Great paper, arguing emergent abilities are only a function of pre training loss and not model/dataset size. ie, if you (inefficiently) overtrain a small model to the loss of GPT4, you'd get all the abilities of GPT4. arxiv.org/abs/2403.15796
New ORPO Colab for Llama-3 8b is out! ORPO combines SFT & DPO into 1 step, so no more 2 step approach! Plus with @UnslothAI, finetuning is 2x faster, uses 80% less VRAM & 4x longer contexts are possible! Thanks to oKatanaaa & At&Dev for making this work! colab.research.google.com/drive/11t4njE3…
one of the most important things I know about deep learning I learned from this paper: "Pretraining Without Attention" this what I found so surprising: these people developed an architecture very different from Transformers called BiGS, spent months and months optimizing it and…
📢#GRaM template for proceedings track is up on our website gram-workshop.github.io . Submit your great ideas in a #ICML like 8-page submission. Accepted papers will be published in PMLR ✨#GRaM proceedings ✨.
Do models need to reason in words to benefit from chain-of-thought tokens? In our experiments, the answer is no! Models can perform on par with CoT using repeated '...' filler tokens. This raises alignment concerns: Using filler, LMs can do hidden reasoning not visible in CoT🧵
Much has been said about many companies’ desire for more compute (as well as data) to train larger foundation models. I think it’s under-appreciated that we have nowhere near enough compute available for inference on foundation models as well. Years ago, when I was leading teams…
In my humble opinion the recent Stream of Search paper (arxiv.org/abs/2404.03683) is truly outstanding. Everyone should give it a thorough read.
I always strongly suggest people to read this work (arxiv.org/abs/2207.10551) by @YiTayML and @m__dehghani when discussing the model architecture. It almost takes up to 50% pages of the literature survey Chapter in my PhD thesis. It is so visionary to study this in 2022. I can…
I always strongly suggest people to read this work (arxiv.org/abs/2207.10551) by @YiTayML and @m__dehghani when discussing the model architecture. It almost takes up to 50% pages of the literature survey Chapter in my PhD thesis. It is so visionary to study this in 2022. I can…
PhysDreamer: Physics-Based Interaction with 3D Objects via Video Generation arxiv.org/abs/2404.13026 Project: physdreamer.github.io Method ⬇️ 1 | 2
Classifier-free guidance has been crucial in recent advancements in content generation, trading diversity for fidelity. However, can you get a better deal in this trade-off? With a simple modification, you can get more diversity while keeping a higher level of details!
Learn2Talk: 3D Talking Face Learns from 2D Talking Face arxiv.org/abs/2404.12888 Project: lkjkjoiuiu.github.io/Learn2Talk/
This #EarthDay you can contribute to improving climate models using #AI with our new kaggle competition (with prizes available!) kaggle.com/competitions/l…
The simplest way to visualize the set of numbers. Utterly useful diagram!❤️ [bit.ly/3hETqNE]
🚨 And we're live: 🎥 Youtube Livestream: tinyurl.com/ys3kdd4k
Toronto Data Workshop summer series starts this Friday. All welcome. Details: rohanalexander.com/events-tdw.htm… Sign up: forms.gle/sXbEixoa1iJR4Q…
Many LLM fine-tuning methods. Unclear what you should use & why? In our new paper, we did an extensive study of on-policy RL, supervised & offline contrastive methods (DPO, IPO) to answer this... 🧵⬇️ On-policy > offline, mode-seeking > mode-covering understanding-rlhf.github.io
Thaytaez @Thaytaez119086
0 Followers 163 FollowingJoyceCotton @69rPFJ2xVfsN6NC
0 Followers 19 FollowingJinyuan (Tobias) @JinyuanWang7
78 Followers 767 Following Research Engineer @NUS. Data scientist @LushairYin Fang @YinFang22900365
518 Followers 517 Following Ph.D. student in CS @ZJU_China. Looking for a post-doc position in AI4Science/LLM/KG. Feel free to reach me if you are interested in my research!_Review_ @Review1266650
26 Followers 2K FollowingTealishoo @tealishoo16978
1 Followers 162 FollowingGram Workshop @GRaM_workshop
32 Followers 182 Following Hi, I am the official account for the first edition of GRaM: Geometry-grounded Representation learning and generative Modeling Workshop at ICML2024Christia Wessell @ChristiaWe36702
86 Followers 5K FollowingSingle Guy Simple Lif.. @singleguysimp1
527 Followers 2K Following I'm a single guy on a mission to live a more simple life through improving my health, finances, and friendships.Agency @AgencyMDR
1K Followers 817 Following Employee Targeted Digital Risk // Personalized Managed Cybersecurity // Security and Compliance for Hyper Growth Companies // YC W22JeanDouglass @7fj7EOfYyrnYFka
0 Followers 140 FollowingNils Lukas @NilsLukas7
159 Followers 297 Following Incoming Assistant Professor @MBZUAI | ML Security & Privacy PhD @UWaterloo | Previous intern @MSFTResearch, @BorealisAIYaeko Nemard @nema_yae
37 Followers 5K FollowingCelia 😏 @CeliaKellie79
3 Followers 566 Following Unstoppablе nуmрho seаrсhing for unbridlеd sаtisfactiоnFreida Vanderbeck @FreidaVand6260
94 Followers 5K FollowingMariko Gorri @GorriMarik69405
62 Followers 5K FollowingAleta @aleta_tran
110 Followers 3K FollowingTessie Miyares @MiyaresTes89108
73 Followers 5K FollowingYu (Bryan) Zhou @yu_bryan_zhou
483 Followers 712 Following Incoming PhD @uclanlp, currently @StanfordSVLTim Kang @TimKang8169
200 Followers 411 Following cs student @uwaterloo, llm enthusiast. @TeraflopAIWei Xiong @weixiong_1
186 Followers 176 Following PhD Student @IllinoisCS, Practice Math for 2.5 YearsMaggie Giulioli @magg_giuli
50 Followers 5K FollowingVladimir Kameñar @KamenarVladimir
358 Followers 382 Following Senior TL at @thalesgroup. CS professor at #UNAL. Published author. All my books are available for free. Programming since the '90s. Opinions are my own.Leia Lean @LeiaLean23287
38 Followers 5K FollowingErika Cardenas @ecardenas300
4K Followers 813 Following @weaviate_io | Interested in vector databases, LLM frameworks, and information retrievalKeri Tafolla @KerTafo
65 Followers 5K FollowingSherilyn Kitson @s_sherily
103 Followers 5K FollowingHarlow Seewald @SeewalHarlo
58 Followers 5K FollowingMarianne Cinotto @CinottoMar46164
96 Followers 5K FollowingEmo_Queen @EmoQueen26830
33 Followers 2K FollowingAleena Detore @AleenaDet
61 Followers 5K FollowingMichaela Mcalpin @McalpinMcal
73 Followers 5K FollowingSaytast @saytast63331
30 Followers 2K FollowingShawanda Daine @shawan_dai
75 Followers 5K FollowingHamid Naderi Yeganeh @naderi_yeganeh
35K Followers 32K Following Research Student @UCL Maths. Mathematical artist. Email: naderiyeganeh at gmail dot comIoannis Kakogeorgiou @IoannisKakogeo1
133 Followers 327 Following I am a Ph.D. candidate at the RS-Lab at the National Technical University of Athens. My research focuses on self-supervised learning and explainable AI.Charlene Stcroix @StcroixCha43410
80 Followers 5K FollowingYixin Wan @yixin_wan_
1K Followers 847 Following PhD student @UCLAComSci | Trustworthy Generative Models | Previously @AmazonScience, @MSFTResearch AsiaLena Shira @lena_shir
84 Followers 5K FollowingLuciana Valk @LucianaVal92400
90 Followers 5K FollowingAgnes Headland @AgnHeadl
72 Followers 5K FollowingAva-rose Strimling @RoStrimli
55 Followers 5K FollowingMatthew Norton @mdnorto
24 Followers 110 FollowingLylah Samayoa @lylah_samay
84 Followers 5K FollowingSlic_kSiren @KsirenSlic63129
37 Followers 2K FollowingScottie Parmele @parm_scott
59 Followers 5K FollowingThelaski @Thelaski129263
37 Followers 2K FollowingBeff Jezos — e/acc .. @BasedBeffJezos
102K Followers 2K Following chief accelerator & founder @ e/acc // thermodynamic priest // Kardashev gradient climber // memetic warlord // building @extropic_aiExtropic @Extropic_AI
30K Followers 28 Following ... . .-.. ..-. -....- .- ... ... . -- -... .-.. .. -. --. / .. -. - . .-.. .-.. .. --. . -. -.-. . / ..-. .-. --- -- / - .... . / ..-. ..- - ..- .-. .Yin Fang @YinFang22900365
518 Followers 517 Following Ph.D. student in CS @ZJU_China. Looking for a post-doc position in AI4Science/LLM/KG. Feel free to reach me if you are interested in my research!Amin Mansouri @m_amin_mansouri
410 Followers 474 Following MSc from @Mila_Quebec. Former intern @ICepfl, @ETH_en, @RIKEN_AIP_ENCanyu Chen @CanyuChen3
842 Followers 2K Following CS Ph.D. student @illinoistech | Truthful, Safe and Responsible LLMs | LLMs Meet Misinformation: https://t.co/up5sEN5r1gHao Liu @haoliuhl
4K Followers 155 Following phd student @berkeley_ai https://t.co/ZNJawlrerS machine learning, neural networks.Wilson Yan @wilson1yan
450 Followers 162 Following PhD student @berkeley_ai. Interested in generative modelsRosmine @rosmine_b
327 Followers 329 Following Senior ML Scientist @ FAANG working on LLMs DM me your ML questionsGram Workshop @GRaM_workshop
32 Followers 182 Following Hi, I am the official account for the first edition of GRaM: Geometry-grounded Representation learning and generative Modeling Workshop at ICML2024Sachin @sacmehtauw
596 Followers 75 Following AI/ML Research Scientist at Apple and Affiliate Assistant Professor at the University of Washington, Seattle. Opinions are my own.SambaNova Systems @SambaNovaAI
4K Followers 713 Following We bring #AI innovations developed in advanced research to organizations around the world. Sign up for updates to stay ahead of AI: https://t.co/bGeeh5JSt0Charles Campbell Robe.. @CCRobertsARK
61K Followers 3K Following Chief Investment Strategist for @ARKInvest. Dr.; systems bio; angel; cofounder of unicorn genomics x AI company Freenome. Disclosure: https://t.co/EueoFlWYzaCathie Wood @CathieDWood
1.7M Followers 400 Following Founder, CEO and CIO @ARKinvest. Thematic portfolio manager for disruptive innovation, mom, economist, and women's advocate. Disclosure: https://t.co/chxRD4oWOdAlex Havrilla @Dahoas1
1K Followers 503 Following Georgia Tech ML Researcher studying neural network learning theory and LLMs for mathematical reasoning. Intern at FAIR, MSFT Research. Co-founder of CarperAI.Nils Lukas @NilsLukas7
159 Followers 297 Following Incoming Assistant Professor @MBZUAI | ML Security & Privacy PhD @UWaterloo | Previous intern @MSFTResearch, @BorealisAIKristian Lum @KLdivergence
22K Followers 1K Following Research Scientist at Google DeepMind | @FAccTConference OG | Past Twitter META, @hrdag & UPenn, UChicago faculty |David @DavidSHolz
54K Followers 5K Following founder @midjourney, prev founder leap motion, nasa, max planckChitwan Saharia @Chitwan_Saharia
3K Followers 289 Following @ideogram_ai Past: Sr. Research Scientist @GoogleAI || B. Tech, CSE, @IITBombayJonathan Ho @hojonathanho
4K Followers 152 FollowingMrNeRF @janusch_patas
4K Followers 427 Following When your splat are so on point, even the pixels are getting jealous | Sharing news around 3DGS progress and improvements | Mostly technical tweets and threadsNicolas DUFOUR @nico_dufour
137 Followers 387 Following PhD student at IMAGINE (ENPC) and GeoVic (Ecole Polytechnique). Working on image generation.Steven Bird @StevenBird
2K Followers 567 Following conducting social and technological experiments in the future evolution of the world's languages... member of @cdu_tell (he/they)Bytez @Bytez
94 Followers 124 Following Production-ready open source AI. Discover, demo, and deploy open source AI models on Bytez. Let's accelerate.Yu (Bryan) Zhou @yu_bryan_zhou
483 Followers 712 Following Incoming PhD @uclanlp, currently @StanfordSVLJindong Wang @jd92wang
3K Followers 410 Following Senior Researcher at @MSFTResearch. Robust machine learning, transfer learning, OOD generalization. Recently: large language models. https://t.co/fUsOEtxAW7Tim Kang @TimKang8169
200 Followers 411 Following cs student @uwaterloo, llm enthusiast. @TeraflopAINeel Nanda @NeelNanda5
13K Followers 89 Following Mechanistic Interpretability lead @DeepMind. Formerly @AnthropicAI, independent. In this to reduce AI X-risk. Neural networks can be understood, let's go do it!Yihe Deng @Yihe__Deng
2K Followers 1K Following CS PhD student @UCLA | Prev. Applied Scientist Intern @AWS | LLM, Multi-modal learningMatt Shumer @mattshumer_
51K Followers 1K Following CEO @HyperWriteAI, @OthersideAI - I make AIs do the impossible.Groq Inc @GroqInc
46K Followers 470 Following Creator of the LPU™ Inference Engine, providing the fastest speed for AI applications, designed & engineered in N. America https://t.co/DsEqVAC5DpMindBranches @MindBranches
11K Followers 1K Following AI enhanced diagrams to help you understand complex concepts quickly.Mufan (Bill) Li @mufan_li
804 Followers 492 Following Postdoc @Princeton ORFE | Prev: PhD @UofTStatSci @VectorInstVladimir Kameñar @KamenarVladimir
358 Followers 382 Following Senior TL at @thalesgroup. CS professor at #UNAL. Published author. All my books are available for free. Programming since the '90s. Opinions are my own.Hao Zhang @haozhangml
3K Followers 263 Following Asst. Prof. @HDSIUCSD and @ucsd_cse running @haoailab. Cofounder and runs @lmsysorg. 20% with @SnowflakeDBBeidi Chen @BeidiChen
6K Followers 343 Following Asst. Prof @CarnegieMellon, Visiting Researcher @Meta, Postdoc @Stanford, Ph.D. @RiceUniversity, Large-Scale ML, a fan of Dota2.Wenhan Xiong @XiongWenhan
800 Followers 566 Following #NLProc Researcher #llama. Making some camelids out of electricity @AIatMetaXuezhe Ma (Max) @MaxMa1987
1K Followers 350 Following Research Lead @USC_ISI and Research Assistant Professor @CSatUSC PhD at CMU ML/NLP @LTIatCMU @CarnegieMellonErika Cardenas @ecardenas300
4K Followers 813 Following @weaviate_io | Interested in vector databases, LLM frameworks, and information retrievalKirill Neklyudov @k_neklyudov
678 Followers 254 FollowingUCLA @UCLA
258K Followers 241 Following The official account for the #1 public university in the nation 7 years in a row. Dedicated to research, education and service.Siyan Zhao @siyan_zhao
780 Followers 486 Following CS PhD student @UCLA | Interested in decision making, LLMs, generative models | Bachelors @UofT EngSciPortal @_portal_
306 Followers 63 Following Home of the TechBio community. LoGG/M2D2/CARE reading groups, blogs, events, and more.Ross Taylor @rosstaylor90
6K Followers 876 Following Something new 🥷. Previously: @paperswithcode, reasoning lead @metaai, Galactica LLM lead, Atlas ML (acq by Meta)arxiv.org/abs/2404.15702 Doesn't seem to perform incredibly, but they have a lot of neat details about the training pipeline
Great paper, arguing emergent abilities are only a function of pre training loss and not model/dataset size. ie, if you (inefficiently) overtrain a small model to the loss of GPT4, you'd get all the abilities of GPT4. arxiv.org/abs/2403.15796
New ORPO Colab for Llama-3 8b is out! ORPO combines SFT & DPO into 1 step, so no more 2 step approach! Plus with @UnslothAI, finetuning is 2x faster, uses 80% less VRAM & 4x longer contexts are possible! Thanks to oKatanaaa & At&Dev for making this work! colab.research.google.com/drive/11t4njE3…
one of the most important things I know about deep learning I learned from this paper: "Pretraining Without Attention" this what I found so surprising: these people developed an architecture very different from Transformers called BiGS, spent months and months optimizing it and…
@WenhuChen It depends on personality, not paradigm. Some people (me) can’t focus on more than one project at once and if they try to multitask they do everything badly. Other people get bored doing one thing, and need to have multiple projects to cycle between. Just do what’s best for you…
📢#GRaM template for proceedings track is up on our website gram-workshop.github.io . Submit your great ideas in a #ICML like 8-page submission. Accepted papers will be published in PMLR ✨#GRaM proceedings ✨.
Do models need to reason in words to benefit from chain-of-thought tokens? In our experiments, the answer is no! Models can perform on par with CoT using repeated '...' filler tokens. This raises alignment concerns: Using filler, LMs can do hidden reasoning not visible in CoT🧵
Much has been said about many companies’ desire for more compute (as well as data) to train larger foundation models. I think it’s under-appreciated that we have nowhere near enough compute available for inference on foundation models as well. Years ago, when I was leading teams…
RESEARCH OPPORTUNITY ALERT. If you're interested in synthetic data, we're recruiting for a Research Scholar to collaborate with Cohere For AI for a 6-month internship. Must be available full-time starting ASAP. DM if you're interested 🥰.
👏 PhD candidate Nils Lukas has received the 2024 Mathematics Doctoral Prize’s top honour. As a first-place recipient, he will receive $1,500 and is nominated for the university-wide Governor General’s Gold Medal. Congrats, Nils! 👏 cs.uwaterloo.ca/news/nils-luka…
PhysDreamer: Physics-Based Interaction with 3D Objects via Video Generation arxiv.org/abs/2404.13026 Project: physdreamer.github.io Method ⬇️ 1 | 2
Learn2Talk: 3D Talking Face Learns from 2D Talking Face arxiv.org/abs/2404.12888 Project: lkjkjoiuiu.github.io/Learn2Talk/
This #EarthDay you can contribute to improving climate models using #AI with our new kaggle competition (with prizes available!) kaggle.com/competitions/l…
The simplest way to visualize the set of numbers. Utterly useful diagram!❤️ [bit.ly/3hETqNE]
🚨 And we're live: 🎥 Youtube Livestream: tinyurl.com/ys3kdd4k
Toronto Data Workshop summer series starts this Friday. All welcome. Details: rohanalexander.com/events-tdw.htm… Sign up: forms.gle/sXbEixoa1iJR4Q…
We got a @Google Research Scholar Award for Planning with LLMs 🎉! Thank you @GoogleAI. Excited to go back to my planning roots and build useful planning tools with LLMs. research.google/programs-and-e…
Our team in FAIR (at Meta) is hiring researchers (RS & PostDoc) to work on the broad topics of text and multimodal LLMs. Location: NY, Seattle or Menlo Park for RS, and Seattle for PostDocs. PostDoc: metacareers.com/jobs/968496244… Research Scientist, AI (PhD): metacareers.com/jobs/752169417…