Yi Lin Sung @yilin_sung
CS PhD student @unccs @uncnlp | Previously intern @MetaAI @MSFTResearch | Multi-modal DL, Efficient fine-tuning. ylsung.github.io Chapel Hill, NC Joined March 2013-
Tweets251
-
Followers523
-
Following730
-
Likes469
🚨 We have postdoc openings at UNC 🙂 Exciting+diverse NLP/CV/ML topics**, freedom to create research agenda, competitive funding, very strong students, many collabs w/ other faculty & universities+companies, superb quality of life/weather. Please apply + help spread the word…
Can we design an efficient & versatile framework to reuse+adapt existing pretrained ControlNets to accurately guide any video/image diffusion model and support diverse controls? 🚨 Introducing Ctrl-Adapter: ➡️ Flexible Compatibility: Adapts any pretrained ControlNet…
It was such a pleasure to appear on this podcast (which has literally hundreds of episodes with great AI folks)! Thanks to @samcharrington for a great conversation on the topics below👇
It was such a pleasure to appear on this podcast (which has literally hundreds of episodes with great AI folks)! Thanks to @samcharrington for a great conversation on the topics below👇
🚨 Introducing SegNext, our #CVPR2024 project, combining the best of specialist and generalist designs for interactive segmentation! ➡️ Low Latency ➡️ High Quality ➡️ Diverse Prompts Recent interactive segmentation methods are taking either one of two approaches: (1)…
🚨 Introducing SegNext, our #CVPR2024 project, combining the best of specialist and generalist designs for interactive segmentation! ➡️ Low Latency ➡️ High Quality ➡️ Diverse Prompts Recent interactive segmentation methods are taking either one of two approaches: (1)…
Gemini 1.5 Pro - A highly capable multimodal model with a 10M token context length Today we are releasing the first demonstrations of the capabilities of the Gemini 1.5 series, with the Gemini 1.5 Pro model. One of the key differentiators of this model is its incredibly long…
Stability AI presents SD3-Turbo Fast High-Resolution Image Synthesis with Latent Adversarial Diffusion Distillation Diffusion models are the main driver of progress in image and video synthesis, but suffer from slow inference speed. Distillation methods, like the
🚨 EnvGen --> LLM iteratively decides+generates visual/embodied environments that are most effective to let the RL game agent play+learn in --> so as to automatically target/focus on progressively/adaptively improving the agent's weaker skills --> and hence very efficient…
🚨 EnvGen --> LLM iteratively decides+generates visual/embodied environments that are most effective to let the RL game agent play+learn in --> so as to automatically target/focus on progressively/adaptively improving the agent's weaker skills --> and hence very efficient…
Can we adaptively generate training environments with LLMs to help small embodied RL game agents learn useful skills that they are weak at? 🤔 👉 Check out EnvGen, an effective+efficient framework in which an LLM progressively generates and adapts training environments based on…
🎉Our work ADaPT on enabling LLM agents to dynamically “adapt” to task complexity & LLM capabilities via recursive decomposition is accepted as #NAACL2024 findings!😄 Many thanks to @alkoller M Hartmann, P Clark, @Ashish_S_AI @mohitban47 @tusharkhot @ai2_aristo @allen_ai @uncnlp
🎉Our work ADaPT on enabling LLM agents to dynamically “adapt” to task complexity & LLM capabilities via recursive decomposition is accepted as #NAACL2024 findings!😄 Many thanks to @alkoller M Hartmann, P Clark, @Ashish_S_AI @mohitban47 @tusharkhot @ai2_aristo @allen_ai @uncnlp
📢New Paper! We introduce EnvGen, a new effective and efficient framework for embodied RL game agents to adaptively learn their weak skills by using LLMs to generate useful/proper training environments! 😆 ▶️A lightweight (4M params) RL agent trained w/ EnvGen outperforms a…
📢New Paper! We introduce EnvGen, a new effective and efficient framework for embodied RL game agents to adaptively learn their weak skills by using LLMs to generate useful/proper training environments! 😆 ▶️A lightweight (4M params) RL agent trained w/ EnvGen outperforms a…
Introducing Evolutionary Model Merge: A new approach bringing us closer to automating foundation model development. We use evolution to find great ways of combining open-source models, building new powerful foundation models with user-specified abilities! sakana.ai/evolutionary-m…
Fine-tuning the LLaMA-2-Chat model may degrade its original capabilities (arxiv.org/abs/2401.03129). But here's a lifeline: Chat Vector (arxiv.org/abs/2310.04799) keeps a chat model's original capability (it also works on Mistral). Recommend to everyone fine-tuning their LLMs.
🚀Check out our #NAACL2024 paper on continual pre-training of language models (LMs)! In the real world, LMs need to unlearn/overwrite outdated information with updated ones. To evaluate the temporal adaptation capabilities of LMs, we introduce the temporally evolving QA…
Some really cool work by our soon-to-be-intern @JialuLi96 link: selma-t2i.github.io
Modern VidQA models are static, operating on fixed training datasets. In contrast, real-world applications demand adaptability to continually changing training domains. We present a parameter-efficient method for continual VidQA learning. arxiv.org/abs/2403.08755 🧵
➡️➡️How to improve multiple skills in Text-to-Image (T2I) models via self-training? 👉 SELMA 👉 -- Generates multiple skill-specific image-text pairs using LLM+T2I. -- Learns skill-specific LoRA experts in parallel (to minimize "knowledge interference" between skills). --…
➡️➡️How to improve multiple skills in Text-to-Image (T2I) models via self-training? 👉 SELMA 👉 -- Generates multiple skill-specific image-text pairs using LLM+T2I. -- Learns skill-specific LoRA experts in parallel (to minimize "knowledge interference" between skills). --…
Introducing SELMA! We teach multiple skills to a T2I model with the following recipes: (1) Automatically generate skill-specific image-text pairs with LLM + T2I (2) Learn skill-specific experts with LoRA in parallel (3) Merge the expert models to obtain a final multi-skill T2I…
Introducing SELMA! We teach multiple skills to a T2I model with the following recipes: (1) Automatically generate skill-specific image-text pairs with LLM + T2I (2) Learn skill-specific experts with LoRA in parallel (3) Merge the expert models to obtain a final multi-skill T2I…
SELMA improves T2I models' text faithfulness and has a better human preference ✅Skill-specific data generation ✅Efficient skill-specific LoRA fine-tuning & Merging ✅Boosting 5 metrics & human eval ✅Weak-to-strong generalization in T2I ✅Comparable performance with GT data
SELMA improves T2I models' text faithfulness and has a better human preference ✅Skill-specific data generation ✅Efficient skill-specific LoRA fine-tuning & Merging ✅Boosting 5 metrics & human eval ✅Weak-to-strong generalization in T2I ✅Comparable performance with GT data
Pointing to an image region should help models focus, but standard VLMs fail to understand visual markers/prompts (e.g., boxes/masks). 🚨Contrastive Region Guidance: Training-free method that increases focus on visual prompts by reducing model priors. arxiv.org/abs/2403.02325 🧵
Are unified VL models consistent across predictions for different tasks on the same image? Thrilled to share our @TmlrOrg paper where we find that VL models show significant cross-task inconsistency in their predictions for the same image across tasks. adymaharana.github.io/cococon/ 🧵
Colin Raffel @colinraffel
30K Followers 655 Following nonbayesian parameterics, sweet lessons, and random birds. Friend of @srush_nlpJaemin Cho @jmin__cho
1K Followers 892 Following PhD student at @UNCCS @UNCNLP Previously at @GoogleAI, @MSFTResearch, @AdobeResearch, @Allen_AI, @official_naver, and @SeoulNatlUniZineng Tang @ZinengTang
1K Followers 569 Following PhD in @Berkeley_ai and @BerkeleyNLP. Previously @UNCNLP and @MSFTResearch.Peter Hase @peterbhase
2K Followers 693 Following Google PhD Fellow at @uncnlp. Interested in interpretable ML, natural language processing, AI Safety, and Effective Altruism.Swarnadeep Saha @swarnaNLP
945 Followers 826 Following @Google PhD Fellow @uncnlp. Formerly @AIatMeta, @SFResearch, @IBMResearch, and @iitdelhi. Gooner.Mohit Bansal @mohitban47
9K Followers 651 Following Parker Distinguished Professor, UNC Chapel Hill (@unc). Director https://t.co/5qlPVgnrlN (@uncnlp). Prev: @Berkeley_AI, @TTIC_Connect @IITKanpur #NLP, #CV, #AI, #MLYichen Jiang @YichenJiang9
652 Followers 472 Following PhD candidate at UNC-Chapel Hill (@uncnlp) | @Apple AI/ML PhD Fellow | Past Intern @Apple @Alexa @MetaAI @MSFTResearch | #NLProc | Working on Compositionality.Min-Hung (Steve) Chen @CMHungSteven
2K Followers 1K Following Senior Research Scientist @NVIDIAAI @NVIDIA | Ex-@Microsoft Azure AI, @MediaTek AI | Ph.D. @GeorgiaTech | Multimodal AI/CV/DL/ML | https://t.co/dKaEzVoTfZLeshem Choshen 🤖�.. @LChoshen
4K Followers 547 Following 🥇 Collaborative LLMs 🥈 Opinionatedly sharing #ML & #NLP 🥉 Propagating us underdogs we owe science an alternative hype @IBMResearch & @MIT_CSAILUNC NLP @uncnlp
3K Followers 388 Following NLP (+ML/AI/CV) research group at UNC ChapelHill (@UNCCS @UNC). Faculty: @mohitban47+@gberta227+@snigdhac25+@shsriva+@tianlongchen4+@huaxiuyaoml + othersLashaun Durland @lasha_durl
31 Followers 5K FollowingAnitraLiverpool @AnitraLive76436
11 Followers 774 FollowingCarri Pinneo @pinne_carr
46 Followers 5K FollowingNithin Sivakumaran @_NithinS
0 Followers 20 FollowingShirleyLocke @gCH49mR9OQ4UZ
2 Followers 92 FollowingAnnTobias @lPt9wj8115ayZ
1 Followers 109 FollowingAaditya ; @Aaditya26082004
532 Followers 7K Following CS'26 • Machine Learning • Open-Source • Web Dev. • Algorithms • Jai Shree Krishna 🦚🪈Kathrine Estorga @EstorgKathrin
84 Followers 5K FollowingTawnya Wydra @tawny_wyd
86 Followers 5K FollowingAllegra Amodeo @AllegraAmo47051
52 Followers 5K FollowingArif Ahmad @arif_ahmad_py
281 Followers 7K Following All things AI, Computer Science and Circuits! Prev. @GoogleAITeemu Summanen @teemusum
195 Followers 3K Following Interested in AI, security, healthcare, and Flutter & Dart.👨🏼💻At X for reading diverse views by professionals and hobbyists.🔬📚🫶Randall @Randall1475089
1 Followers 198 FollowingMatilda Peterson @PetersonMa1571
84 Followers 5K FollowingMN_Noor @noorieBytes
0 Followers 7 FollowingAmartya Banerjee @eigenamartya
28 Followers 641 Following PhD student @unccs | Undergrad @UofMaryland '20 Math + CSSueann Sotto @suea_sot
52 Followers 5K FollowingMadisyn Ude @madisy_ud
45 Followers 5K FollowingMicah Cervetti @MicCervet
31 Followers 5K FollowingKai Wen Cui @cuikaiwen
237 Followers 3K Following CS and Engineering @ Polytechnic University of MilanZaid Khan @codezakh
167 Followers 336 Following @uncnlp with @mohitban47 working on grounded language understanding & reasoning + multimodal agents // researcher @NECLabsAmerica // bs+ms CompE @northeasternLogan Scheidler @LoganScheid
64 Followers 5K FollowingMegan Halasz @MegaHalasz
82 Followers 5K FollowingYajaira Torregrossa @YTorregros41869
36 Followers 5K Followingkunal singh @ikunalsingh7
62 Followers 669 Following Lead AI Researcher https://t.co/z4idFlmggM (T2I), Lead AI Researcher @fractalai Prev: GSoC @CERN, Alumni @IITKgp, Intern @AmiiThinks Diffusion, VLMs, reasoning@LLMAlexander Wan @alexwan55
475 Followers 944 Following CS at Berkeley; @BerkeleyML @BerkeleyNLP; NLP researchEphemeral @Ephemeral862641
4 Followers 46 FollowingYufan Song @YufanSong98
22 Followers 260 FollowingAl Mamun @al_mamun_sardar
276 Followers 3K Following Looking for NLP/LLM related PhD positions | Research Assistant (NLP) at Jahangirnagar University | MSc (CS)Eli Chien @chien_eli
156 Followers 210 Following Postdoc at @GeorgiaTech. Ph.D. from @UofIllinois. Focus on privacy + graph learning. #MachineUnlearning #DifferentialPrivacy #DP #GNNbizika @bizika7
22 Followers 380 FollowingHaokun Liu @liu_haokun
184 Followers 199 Following Some student @ UNC Chapel Hill | Likes: many NLP stuff, empirical study, Wagner, (shamefully) github desktop | he/himandy 徐 @Snn5Ki
34 Followers 108 Following 1.曾在大厂,成熟业务到新项目,新项目从300人起步发展到超过2.5万人 2.大众创业,万众创新浪潮中游泳 3.500强上市公司销售总监Sami Nas 👨⚕�.. @digitalhealthxx
8K Followers 9K Following Senior functional/technical consultant to bring added value via #digitalhealth #ai and #datascience based solutions #MedTwitterFaye Fillare @fill_fa
18 Followers 3K FollowingSahithya Ravi @Sahithya_Ravi
303 Followers 499 Following PhD Student @UBC_CS | @VectorInst I @UBC_NLPSunshine @rachie_baby21
496 Followers 1K Following I can't tolerate a mediocre life, I hope to use my life to create value.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/q2Qoey80GxColin Raffel @colinraffel
30K Followers 655 Following nonbayesian parameterics, sweet lessons, and random birds. Friend of @srush_nlpAndrej Karpathy @karpathy
980K Followers 905 Following 🧑🍳. Previously Director of AI @ Tesla, founding team @ OpenAI, CS231n/PhD @ Stanford. I like to train large deep neural nets 🧠🤖💥AI at Meta @AIatMeta
532K Followers 255 Following Together with the AI community, we are pushing the boundaries of what’s possible through open science to create a more connected world.Yi Tay @YiTayML
29K Followers 97 Following chief scientist / cofounder @RekaAILabs 🫠 past: research scientist @google brain 🤯 currently learning to be a dad 🍼Yann LeCun @ylecun
712K Followers 719 Following Professor at NYU. Chief AI Scientist at Meta. Researcher in AI, Machine Learning, Robotics, etc. ACM Turing Award Laureate.Jaemin Cho @jmin__cho
1K Followers 892 Following PhD student at @UNCCS @UNCNLP Previously at @GoogleAI, @MSFTResearch, @AdobeResearch, @Allen_AI, @official_naver, and @SeoulNatlUniZineng Tang @ZinengTang
1K Followers 569 Following PhD in @Berkeley_ai and @BerkeleyNLP. Previously @UNCNLP and @MSFTResearch.Yoav Artzi @yoavartzi
13K Followers 163 Following Research/prof @cs_cornell + @cornell_tech🚡 / https://t.co/9YnWry7yHs / https://t.co/3VmRSyYm2d / asso. faculty director @arxiv / building https://t.co/f9QkzO5kaCJia-Bin Huang @jbhuang0604
51K Followers 285 Following Associate Professor @umdcs; Part-time Research Scientist @Meta. I like pixels.Xin Eric Wang @xwang_lk
7K Followers 1K Following Multimodal and Embodied AI Researcher / Professor @UCSC. Director of https://t.co/Y4swOBag21. AI for Humanity in the long run. he/himFrançois Chollet @fchollet
470K Followers 769 Following Deep learning @google. Creator of Keras. Author of 'Deep Learning with Python'. Opinions are my own.Jim Fan @DrJimFan
229K Followers 3K Following @NVIDIA Sr. Research Manager & Lead of Embodied AI (GEAR Lab). Creating foundation models for Humanoid Robots & Gaming. @Stanford Ph.D. @OpenAI's first intern.William Wang @WilliamWangNLP
14K Followers 718 Following UCSB NLP Lab + ML Center. https://t.co/6TOnqbk6YT https://t.co/KJYhnav3Et Mellichamp Chair Prof. at UCSB CS. PhD @ CMU SCS. Areas: #NLProc, Machine Learning, AI.Percy Liang @percyliang
49K Followers 408 Following Associate Professor in computer science @Stanford @StanfordHAI @StanfordCRFM @StanfordAILab @stanfordnlp | cofounder @togethercompute | PianistNeurIPS Conference @NeurIPSConf
112K Followers 35 Following New Orleans, Dec 10-16, 23. https://t.co/ga8aOw615g Tweets to this account are not monitored. Please send feedback to [email protected].Akari Asai @AkariAsai
11K Followers 650 Following Ph.D. student @uwcse & @uwnlp. NLP. IBM Ph.D. fellow (2022-2023). Meta student researcher (2023-) . ☕️ 🐕 🏃♀️🧗♀️🍳Zaid Khan @codezakh
167 Followers 336 Following @uncnlp with @mohitban47 working on grounded language understanding & reasoning + multimodal agents // researcher @NECLabsAmerica // bs+ms CompE @northeasternZengyi Qin @qinzytech
1K Followers 178 Following MIT PhD @MIT | Co-founded @myshell_ai | ex. @Stanford @MSFTResearch | Let's do AGI!Ani Kembhavi @anikembhavi
2K Followers 297 Following Senior Director @allen_ai + Affiliate Assoc Prof @UW 📷 : Visual Prog, Unified-IO, BiDAF 🤖 : ProcTHOR, Objaverse, SPOC 🌎 : SATLAS All views my own.John Schulman @johnschulman2
39K Followers 611 Following Cofounder @openai, lead post-training for ChatGPT and the API. Interested in reinforcement learning, alignment, birds, jazz musicChris Olah @ch402
91K Followers 173 Following Reverse engineering neural networks at @AnthropicAI. DMs open! Previously @distillpub, OpenAI Clarity Team, Google Brain. Personal account.Petar Veličković @PetarV_93
30K Followers 555 Following Staff Research Scientist @GoogleDeepMind | Affiliated Lecturer @Cambridge_Uni | Associate @clarehall_cam | GDL Scholar @ELLISforEurope. Monoids. 🇷🇸🇲🇪🇧🇦Tim Rocktäschel @_rockt
29K Followers 1K Following Open-Endedness Team Lead @GoogleDeepMind, Professor of AI @AI_UCL, PI @UCL_DARK, @ELLISforEurope Scholar. ex @MetaAI (FAIR), @CompSciOxford. Opinions my own.Sebastian Riedel (@ri.. @riedelcastro
15K Followers 470 Following Researcher in NLP/ML @deepmind, @ucl_nlp, @[email protected] on MastodonLerrel Pinto @LerrelPinto
5K Followers 181 Following Assistant Professor of CS @nyuniversity. I like robots!Hao Su @haosu_twitr
4K Followers 351 Following Associate Professor @UCSanDiego. Computer Vision, Graphics, Embodied AI, Robotics. Co-Founder of Hillbot Inc.Shuran Song @SongShuran
7K Followers 426 Following Assistant Professor @Stanford University working on #Robotics #AI #ComputerVisionYuke Zhu @yukez
15K Followers 464 Following Assistant Professor @UTCompSci | Co-Leading GEAR @NVIDIAAI | CS PhD @Stanford | Building generalist robot autonomy in the wild | Opinions are my ownArchit Sharma @archit_sharma97
4K Followers 340 Following Final-year CS PhD student @Stanford. Previously, AI Resident @Google Brain, undergraduate @IITKanpur, research intern @MILAMontreal.Stability AI @StabilityAI
190K Followers 31 Following We are building the foundation to activate humanity's potential.Xavier Bresson @xbresson
13K Followers 859 Following Prof @NUSingapore Distinguished Researcher @DiscoverElement #NRF Fellow, #GraphNNs #LLMs #DeepLearningTheory #MolecularMaterialScience #Teaching Opinions my ownAnsh Khurana @AnshKhurana11
2K Followers 656 Following ML @Apple, MS CS @Stanford. Previously, Research @GoogleAI; CS @iitbombay. Views are personal.Fei Xia @xf1280
6K Followers 696 Following Research Scientist at @GoogleDeepMind, Robot Learning, Computer Vision. PhD from @StanfordAILab @StanfordSVL, previously @Tsinghua_Uni. #AGI through EmbodimentMohit Iyyer @MohitIyyer
6K Followers 1K Following assoc. prof at @umasscs, member of @UMass_NLP. i work on natural language processing and deep learningDanfei Xu @danfei_xu
6K Followers 1K Following Assistant Prof. at Georgia Tech @ICatGT, researcher at @NVIDIAAI | Ph.D. @StanfordAILab | Making robots smarterOmar Sanseviero @osanseviero
32K Followers 2K Following Chief Llama Officer @huggingface 🦙 Founder @AI_Learners. Xoogler (SWE @Google Assistant, 20% PM TF Graphics). 100% Hacker Llama🇵🇪🇲🇽Bolei Zhou @zhoubolei
6K Followers 977 Following Assistant Professor at Computer Science Department @UCLAComSci @UCLAengineering @UCLATaco Cohen @TacoCohen
21K Followers 3K Following Deep learner at FAIR. Into codegen, equivariance, generative models. Spent time at Qualcomm, Scyfer (acquired), UvA, Deepmind, OpenAI.Zhiyu Zoey Chen @ZhiyuChen4
1K Followers 303 Following NLP researcher. Postdoc @S3DatCMU. Incoming Assistant Professor @UT_Dallas. PhD @UCSBCS. #NLProc.Stephan Mandt @StephanMandt
2K Followers 556 Following ML Professor @UCIrvine, previously @blei_lab, @Princeton. #GenerativeAI, #Compression, #AI4Science. Program Chair @aistats_conf 2024; General Chair AISTATS 2025Hang Zhao @zhaohang0124
2K Followers 836 Following Asst. Prof@ Tsinghua University, former Scientist@Waymo, MIT PhD’19. Researching on Multimodal Learning and Autonomous Driving, and Robot Learning.Ruohan Gao @RuohanGao1
2K Followers 378 Following Incoming Assistant Professor @umdcs and Research Scientist @RealityLabs. Previously PostDoc @StanfordAILab and Ph.D. @UTCompSci. I teach machines 👀👂🖐️.Sung Ju Hwang @SungJuHwang1
366 Followers 139 Following Associate Professor in the Graduate School of AI & School of Computing @ KAISTSakana AI @SakanaAILabs
19K Followers 0 Following We are a Tokyo-based R&D company on a quest to create a new kind of foundational AI model based on nature-inspired intelligence. https://t.co/LonvHEtlJRYunzhu Li @YunzhuLiYZ
4K Followers 451 Following Assistant Professor of Computer Science @ UIUC @UofIllinois @IllinoisCS, Postdoc from @Stanford @StanfordSVL, PhD from @MIT_CSAIL. #Vision #Robotics #LearningLin Shao @linshaonju
2K Followers 3K Following Assistant Professor in Robotics @NUS | Ph.D. @Stanford | Opinions are my ownXiang Yue @xiangyue96
2K Followers 434 Following Postdoc @LTIatCMU. PhD from Ohio State @osunlp. Training & evaluating foundation models. Pushing the boundaries of AI🤖. Previously @MSFTResearch.Huazhe Harry Xu @HarryXu12
2K Followers 895 Following Hi, I like reinforcement learning, robots, and video games:) I am an amateur pianist. Assistant Prof at Tsinghua; Postdoc at Stanford; Ph.D. at BerkeleyRobert Stojnic @rbstojnic
3K Followers 488 Following Open source AI. ⌛Past: Llama 2 and Llama 3 technical leadership at Meta AI, Papers with Code co-creator.Han Wang @HanWang98
76 Followers 328 Following PhD student at UNC-Chapel Hill (@unc @unccs @uncnlp); Formerly Intern @MSFTResearch @NlpWestlake. RT & like ≠ endorsements. Views are my own. He/himDavid Ifeoluwa Adelan.. @davlanade
2K Followers 1K Following @DeepMind Academic Fellow @uclcs, incoming assistant Professor @mcgillu, Canada CIFAR AI Chair @CIFAR_News | interested in multilingual NLP | Disciple of JesusAl Mamun @al_mamun_sardar
276 Followers 3K Following Looking for NLP/LLM related PhD positions | Research Assistant (NLP) at Jahangirnagar University | MSc (CS)Eli Chien @chien_eli
156 Followers 210 Following Postdoc at @GeorgiaTech. Ph.D. from @UofIllinois. Focus on privacy + graph learning. #MachineUnlearning #DifferentialPrivacy #DP #GNNXingyao Wang @xingyaow_
896 Followers 938 Following PhD student @IllinoisCS | BS @UMichCSE ('22) | Ex Intern @GoogleAI @Microsoft | Natural Language Processing | OpenDevin Core ContributorAllan Zhou @AllanZhou17
1K Followers 447 Following Final-year AI PhD student @Stanford. NN architecture design, learned optimizers, and hparam optimization.Chujie Zheng @ChujieZheng
520 Followers 496 Following LLM alignment and safety #LLMs | Visiting Scholar @CS_UCLA | PhD student @TsinghuaCoAI | he/him/hisMajeed Kazemi @MajeedKazemi
1K Followers 2K Following PhD student in CS @UofT with @ToviGrossman HCI + Computing Education + Coding / Creativity Support Tools Prev: @MSFTResearch + MSc @HCIL_UMD with @JonFroehlichChaitanya K. Joshi @chaitjo
6K Followers 2K Following PhD student at University of Cambridge @Cambridge_CL. Interested in Graph & Geometric Deep Learning + Biomolecule modelling & design. Organising @LoGConference.Sara Hooker @sarahookr
39K Followers 7K Following I lead @CohereForAI. Formerly Research @Google Brain @GoogleDeepmind. ML Efficiency at scale, LLMs, @trustworthy_ml. Changing spaces where breakthroughs happen.Allie K. Miller @alliekmiller
49K Followers 2K Following #1 Most Followed Voice in AI Business (1.5M followers). Nat’l AAAS Ambassador. Former Amazon, IBM. Fortune 500 and startup AI advisor, public speaker.🚨 We have postdoc openings at UNC 🙂 Exciting+diverse NLP/CV/ML topics**, freedom to create research agenda, competitive funding, very strong students, many collabs w/ other faculty & universities+companies, superb quality of life/weather. Please apply + help spread the word…
We have a new preprint out - your language model is not a reward, it’s a Q function! 1. The likelihood of the preferred answer must go down - it’s a policy divergence 2. MCTS guided decoding on language is equivalent to likelihood search on DPO 3. DPO learns credit assignment
It’s here! Meet Llama 3, our latest generation of models that is setting a new standard for state-of-the art performance and efficiency for openly available LLMs. Key highlights • 8B and 70B parameter openly available pre-trained and fine-tuned models. • Trained on more…
Introducing 🔌Ctrl-Adapter🔌, our plug-and-play framework that reuses any existing ControlNet for any video/image diffusion model! ➡️ Plug-and-Play: Adapts any (smaller) ControlNet to any (larger) image (e.g., SDXL) or video diffusion (e.g., Stable Video Diffusion, I2VGen-XL,…
Can we design an efficient & versatile framework to reuse+adapt existing pretrained ControlNets to accurately guide any video/image diffusion model and support diverse controls? 🚨 Introducing Ctrl-Adapter: ➡️ Flexible Compatibility: Adapts any pretrained ControlNet…
🎉Excited to share our new Ctrl-Adapter framework which can efficiently adapt any existing pretrained ControlNet to any video/image diffusion model! In <10 hours of GPU training, Ctrl-Adapter can outperform baselines/methods. Ctrl-Adapter also enables a broad range of…
Can we design an efficient & versatile framework to reuse+adapt existing pretrained ControlNets to accurately guide any video/image diffusion model and support diverse controls? 🚨 Introducing Ctrl-Adapter: ➡️ Flexible Compatibility: Adapts any pretrained ControlNet…
Can we design an efficient & versatile framework to reuse+adapt existing pretrained ControlNets to accurately guide any video/image diffusion model and support diverse controls? 🚨 Introducing Ctrl-Adapter: ➡️ Flexible Compatibility: Adapts any pretrained ControlNet…
It was such a pleasure to appear on this podcast (which has literally hundreds of episodes with great AI folks)! Thanks to @samcharrington for a great conversation on the topics below👇
Today we're joined by @peterbhase from @uncnlp to discuss mechanistic interpretability, scalable oversight, and how matrix probing techniques begin to illuminate the "black box" of large neural networks. 🎧 / 🎥 Listen or watch the full episode at: twimlai.com/go/679. 📖…
Today we're joined by @peterbhase from @uncnlp to discuss mechanistic interpretability, scalable oversight, and how matrix probing techniques begin to illuminate the "black box" of large neural networks. 🎧 / 🎥 Listen or watch the full episode at: twimlai.com/go/679. 📖…
The next chapter about transformers is up on YouTube, digging into the attention mechanism: youtu.be/eMlx5fFNoYc The model works with vectors representing tokens (think words), and this is the mechanism that allows those vectors to take in meaning from context.
Introducing our #CVPR2024 paper 🔥 SegNext 🔥 We propose SegNext as a next-generation interactive image segmentation method to support all the following useful features: ➡️ low latency ➡️ high quality ➡️ diverse prompts arxiv.org/abs/2404.00741
🚨 Introducing SegNext, our #CVPR2024 project, combining the best of specialist and generalist designs for interactive segmentation! ➡️ Low Latency ➡️ High Quality ➡️ Diverse Prompts Recent interactive segmentation methods are taking either one of two approaches: (1)…
Introducing our #CVPR2024 paper 🔥 SegNext 🔥 We propose SegNext as a next-generation interactive image segmentation method to support all the following useful features: ➡️ low latency ➡️ high quality ➡️ diverse prompts arxiv.org/abs/2404.00741
🔊Check out our preprint, AVSiam 🤝. We use a single shared ViT to process audio and visual inputs, improving its parameter efficiency⚡, reducing the GPU memory footprint💻, and allowing us to scale📈 to larger datasets and model sizes. arxiv.org/abs/2403.19638 w. @gberta227 👇
``Siamese Vision Transformers are Scalable Audio-visual Learners,'' Yan-Bo Lin, Gedas Bertasius, ift.tt/feFJkwQ
Can we improve narrative closure with bookending, i.e., relating the last sentence back to the first sentence? We present RENarGen, a paradigm for both LMs and LLMs that generates narratives with related endpoint sentences. Accepted at #NAACL2024
Exciting new work with @AnnelieseB_ and @zhaochaocs on generating stories that provide narrative closure. It will be presented at NAACL'24 @UNCResearch @unccs @uncnlp
Can we improve narrative closure with bookending, i.e., relating the last sentence back to the first sentence? We present RENarGen, a paradigm for both LMs and LLMs that generates narratives with related endpoint sentences. Accepted at #NAACL2024
That's the exact opposite IMO! $10M to train a GPT3.5 level model whereas it probably cost OAI at least 10-20x more just a year or two ago. The more we improve as a field thanks to open-source, the cheaper & more efficient it gets to produce the same capabilities. Let's go…
Our team at Google DeepMind has a full-time Research Scientist position available at our Mountain View site. Minimum qualification: PhD in ML/NLP. Please email me with: your CV and Google Scholar link; a brief description of the impactful work you have done; and what you aim…
Introduce 3D-VLA, a 3D Generative World Model! Humans use mental models to predict and plan for the future. Similarly, 3D-VLA achieves this by linking 3D perception, future prediction, and action executions through a generative world model. vis-www.cs.umass.edu/3dvla
Should we acquire Stability and open-source SD3?
Our paper, "The LLM Surgeon," accepted at ICLR 2024, achieves SOTA in LLM pruning in all unstructured, semi-structured, and the most challenging but most effective structured pruning that removes entire matrix rows/columns. Happy to share that code is now publicly available.…
The LLM Surgeon paper page: huggingface.co/papers/2312.17… State-of-the-art language models are becoming increasingly large in an effort to achieve the highest performance on large corpora of available textual data. However, the sheer size of the Transformer architectures makes it…