Jacob Andreas @jacobandreas
Teaching computers to read. Assoc. prof @MITEECS / @MIT_CSAIL (he/him). https://t.co/5kCnXHjtlY https://t.co/2A3qF5vdJw Cambridge, MA Joined March 2007-
Tweets3K
-
Followers13K
-
Following955
-
Likes1K
MAIA (A Multimodal Automated Interpretability Agent) is here! 🧵 📝New paper: arxiv.org/abs/2404.14394 🌐Website: …imodal-interpretability.csail.mit.edu/maia/ Agents like MAIA advance automated interpretation of AI systems from one-shot feature description into an interactive regime where hypotheses…
A Multimodal Automated Interpretability Agent This paper describes MAIA, a Multimodal Automated Interpretability Agent. MAIA is a system that uses neural models to automate neural model understanding tasks like feature interpretation and failure mode discovery. It equips a
Thanks @_akhaliq! MAIA is next in our line of work on Interpretability Agents that interrogate other models, and the functions of their components, using iterative experimentation. Project page: …imodal-interpretability.csail.mit.edu/maia/ w/ @TamarRottShaham @f_x_wang @AchyutaBot @evanqed…
Thanks @_akhaliq! MAIA is next in our line of work on Interpretability Agents that interrogate other models, and the functions of their components, using iterative experimentation. Project page: …imodal-interpretability.csail.mit.edu/maia/ w/ @TamarRottShaham @f_x_wang @AchyutaBot @evanqed…
We should be smarter than just scaling! We should create data-efficient algs. Humans are great at this, algs should learn from humans. This is what I have been working on (t.ly/KB818). This is also what our BabyLM is about (babylm.github.io)!
We should be smarter than just scaling! We should create data-efficient algs. Humans are great at this, algs should learn from humans. This is what I have been working on (t.ly/KB818). This is also what our BabyLM is about (babylm.github.io)!
As many have said already, this is a terrible idea. Invest in the inclusion of the Global South instead, that's what's missing. We don't need to entice adolescents to write papers. #NeurIPS2024
As many have said already, this is a terrible idea. Invest in the inclusion of the Global South instead, that's what's missing. We don't need to entice adolescents to write papers. #NeurIPS2024
Two papers! Can visual grounding help LMs learn more efficiently? 1. We show that algs like CLIP don't learn language better (t.ly/eQHA9) 2. We then propose a new one, LexiContrastive Grounding, which does! (t.ly/KB818) Code: t.ly/C0wu- 🧵
Some very cool recent work from @ChengxuZhuang on visually grounded language learning. Surprisingly, standard text/image repr learning gives great image reprs but doesn't change behavior much on language tasks---how can we use image data for better learning of *language*?
Some very cool recent work from @ChengxuZhuang on visually grounded language learning. Surprisingly, standard text/image repr learning gives great image reprs but doesn't change behavior much on language tasks---how can we use image data for better learning of *language*?
Our second BabyLM Challenge is here! This year, we feature a vision-language track. You can also bring your own data as long as it has less than 100M words! See babylm.github.io for more info. Let's develop baby-like language learning algs!
Our second BabyLM Challenge is here! This year, we feature a vision-language track. You can also bring your own data as long as it has less than 100M words! See babylm.github.io for more info. Let's develop baby-like language learning algs!
New work on the Battleship Game accepted to CogSci '24! ⚓️🧠 How do people pose informative, grounded questions in uncertain environments? And how can we build machines that ask human-like questions? arxiv.org/abs/2402.19471 🧵 (1 / n)
🧙🍪🧙♀️ I'm hiring a postdoc with @sebschu to start in Fall 2024! We are looking for someone with experience in EITHER: (1) building systems that use language models as a core component to solve complex tasks, or (2) leading human annotation/behavioral experiments.
📢 I'm recruiting a postdoc at UMD 🗓️ Starting in 2024 🖥️🙋 Topics related to human-AI interaction 🔜 Apply by May 31, but earlier if possible! 🤑 $75-$80k+benefits, unencumbered (ie not "big project") funding Link to full post and form for applications in thread. >
We're looking for a brilliant postdoc to work with @roger_p_levy @nidhi_s91 and me on an exciting new project at the intersection of computational cognition, language, and motor control! Please share with anyone who might be interested. More info here: academicjobsonline.org/ajo/jobs/27294
I am hoping to hire a postdoc who would start in Fall 2024. If you are interested in the intersection of linguistics, cognitive science, and AI, I encourage you to apply! Please see this link for details: rtmccoy.com/prospective_po…
Looking for 1-2 FT RAs to start this summer. It's a 2- year position ideal for folks graduating college this spring who are looking for more research experience before applying to PhD programs in cog sci / neurosci. For full consideration, apply by Mar 15: tinyurl.com/ywxt24x3
Turns out that @alsuhr's good ol' fashioned (2017!) NLVR remains pretty challenging for SOTA multimodal LLMs ¯\_(ツ)_/¯ New technical report by @anne_youw Particularly striking given the tiny vocabulary size and the simple synthetic images. Why? Not completely sure, but ...
Turns out that @alsuhr's good ol' fashioned (2017!) NLVR remains pretty challenging for SOTA multimodal LLMs ¯\_(ツ)_/¯ New technical report by @anne_youw Particularly striking given the tiny vocabulary size and the simple synthetic images. Why? Not completely sure, but ... https://t.co/8eodcREcK4
How can we build AI assistants that *reliably* follow our instructions, even when they're ambiguous? @Lance_Ying42 & I introduce CLIPS: A Bayesian arch. combining inverse planning w LLMs that *pragmatically* infers human goals from actions & language, then provides assistance!
Ever wondered how finetuning boosts a language model's performance? Our ICLR24 paper (w @TamarRottShaham @Tal_Ha535 @boknilev @davidbau) unveils the secret: fine-tuning enhances, rather than fundamentally alters, the existing mechanisms of original model. finetuning.baulab.info
(((ل()(ل() 'yoav))).. @yoavgo
46K Followers 2K FollowingPercy Liang @percyliang
49K Followers 408 Following Associate Professor in computer science @Stanford @StanfordHAI @StanfordCRFM @StanfordAILab @stanfordnlp | cofounder @togethercompute | PianistKyunghyun Cho @kchonyc
61K Followers 2K Following a combination of a mediocre scientist, a mediocre manager, a mediocre advisor & a mediocre PC at @nyuniversity (@CILVRatNYU) & @genentech (@PrescientDesign).Sam Bowman @sleepinyourhat
35K Followers 3K Following AI alignment + LLMs at NYU & Anthropic. Views not employers'. No relation to @s8mb. I think you should join @givingwhatwecan.Sasha Rush @srush_nlp
52K Followers 464 Following Professor, Programmer in NYC. Cornell Tech, Hugging Face 🤗 https://t.co/cZl0wTfqGzYoav 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/f9QkzO5kaCAkari Asai @AkariAsai
11K Followers 650 Following Ph.D. student @uwcse & @uwnlp. NLP. IBM Ph.D. fellow (2022-2023). Meta student researcher (2023-) . ☕️ 🐕 🏃♀️🧗♀️🍳Tal Linzen @tallinzen
16K Followers 893 Following Professor @nyuling and @NYUDataScience, research scientist @GoogleAIGraham Neubig @gneubig
31K Followers 586 Following Associate professor at CMU, studying natural language processing and machine learning.Natasha Jaques @natashajaques
25K Followers 1K Following Senior Research Scientist at @GoogleAI and Assistant Professor @uwcse. Social Reinforcement Learning in multi-agent and human-AI interactions. PhD from @MIT.Felix Hill @FelixHill84
9K Followers 777 Following Research Scientist, Deepmind I try to think hard about everything I tweet, esp on 90s football and 80s music None of my opinions are really someone else'sEdward Grefenstette @egrefen
36K Followers 774 Following FR/US/GB AI/ML Person, Director of Research at @GoogleDeepMind, Honorary Professor at @UCL_DARK, @ELLISforEurope Fellow. All posts are personal.Naomi Saphra @nsaphra
7K Followers 1K Following Waiting on a robot body. ML/NLP. All opinions are universal and held by both employers and family. Same username on every lifeboat off this sinking ship.Danish Pruthi @danish037
7K Followers 628 Following Faculty at Indian Institute of Science, Bangalore. PhD from @LTIatCMU.Tim Dettmers @Tim_Dettmers
29K Followers 820 Following PhD Student at @UW. I blog about deep learning and PhD life at https://t.co/Y78KDJJFE7.Zachary Lipton @zacharylipton
59K Followers 2K Following Professor: CMU/@acmi_lab, CTO / CSO: @AbridgeHQ, Creator: @d2l_ai & https://t.co/QQt98VNLUp, Relapsing 🎷Kayo Yin @kayo_yin
8K Followers 556 Following PhD student @berkeley_ai @berkeleynlp working on interpretability and signed languages. Former @msftresearch @deepmind @carnegiemellon @polytechnique. 🇫🇷🇯🇵Christopher Potts @ChrisGPotts
11K Followers 620 Following Stanford Professor of Linguistics and, by courtesy, of Computer Science, and member of @stanfordnlp and @StanfordAILab. He/Him/His.Viviana @Viviana75842443
2 Followers 161 FollowingElectronicsseeker @libertarian108
7 Followers 913 FollowingAlexander Morosow @alex5m6
3 Followers 35 Following Head of Creative Engineering & Software Architect @refikanadol studio | @datalandmuseum | simplify omnidirectional motionLara Swara @swaralaraa
661 Followers 648 Following The beauty of an independent woman cannot be compared with the frivolous things out there.Harshal Nandigramwar @hnanacc
343 Followers 247 Following ai @intel labs, prev: ai @cariad_tech, masters @Uni_Stuttgart, building @todackcom, @themelioaiNoman Tanveer @NomanTa98551465
2 Followers 96 Following Interested in Deep Generative Models and Multimodal research!Bharat Kumar Seervi�.. @BharatKumarSee1
89 Followers 641 Following no left ,no right , only rational tech enthusiasts computer science 🖥️ and neuroscienceFer @otferdam
3 Followers 437 Following 29 | Lingüista computacional | UBA-Puán💚 | Fantasía épica+weird fiction 📚🎮🎭🏳️🌈Mengdi Wang @MengdiWang10
1K Followers 265 Following Princeton professor in AIML, optimization and data science. Program Chair @ICLR2023. Formerly @MIT @GoogleDeepmind @TsinghuaDeependu @deependu__
65 Followers 325 Following Transformers & AWS | Exploring Reinforcement Learning | GitHub: https://t.co/6xNi6IHCzOdev potatopotato @devpotatopotato
2 Followers 136 Following CS student in Seoul National University. Passionate about AGI.Archer Wang @ArcherWang6
66 Followers 139 Following incoming phd student @MITEECS prev B.S. @MIT_Physics / @tryramp @MetaAIAshant Chalasani @ashant
88 Followers 140 FollowingXinyi Wang @XinyiWang98
797 Followers 299 Following UC Santa Barbara CS PhD student working on ML/NLP🍃 @abstract_ing
0 Followers 232 FollowingBilly Porter @porterbilly57
0 Followers 25 Following Research Engineer @Google Labs | NYC | Notre Dame | Building LLMs that read, write, and understand codeविष्णुद.. @visnu_daas
5 Followers 75 FollowingYesu Mweusi @MweusiYesu
0 Followers 28 FollowingWenzhao Qiu @WenzhaoQiu
6 Followers 138 FollowingAndrewRayHerndon @mrHerndon
39 Followers 225 Following Researcher/Designer/Producer https://t.co/zFqb2Kf1l9 Voice Technology Empathetic Research UX Design Tang Soo Do/Shudokan Karate +Karaoke are my main things...Anurag Mishra @anuragm75160136
112 Followers 801 Following Building Scalable AI Applications | Senior Data Scientist @ EY | CSE Btech @ NIT MN | Linkedin: https://t.co/pCmSV6FmOeArhant Chaterjee @ArhantC69420
105 Followers 832 FollowingFalalu Ibrahim Lawan @falalu247
99 Followers 1K Following Educator, STEM ambassador, mentor, passionate about learning and codingNick Mumero @nickdee96
131 Followers 1K Following Cofounder at Continuum Ads. Focusing on NLP, Simulation Modelling and Optimization.Abdul hai mondal @hai_mondal94903
1 Followers 16 FollowingTatsuya Aoyama @t_aoyam
20 Followers 72 Following Ph.D. student in Computational Linguistics @ Georgetown UniversitySamadeep @samadeepviews
106 Followers 1K Following Incoming Software Engineering Intern @GoogleIndia Computer Science UndergradAlice Baird @Aliceebaird
864 Followers 209 Following AI research scientist @hume_ai, PhD from @uni__augsburg - affective computing, computational paralinguistics, wellbeing.CompoSecure @CompoSecure
2K Followers 3K Following CompoSecure, Inc. (Nasdaq: CMPO) is a leader in metal payment cards, security, and authentication solutions.Simon Dobnik @SimonDobnik
119 Followers 287 Following Professor at University of Gothenburg, Sweden. NLP researcher and lecturer.Ji-An Li @Ji_An_Li
159 Followers 689 Following NGP student at UCSD | Computational neuroscience | Neural networks | Marcelo Mattar Lab | Marcus Benna LabJustin Wong @justinwong8314
85 Followers 171 Following CS PhD Student at UC Berkeley advised by Joseph Gonzalez and Sanjit Seshia.deepak @deepakgujraniya
31 Followers 171 FollowingChristian Moya Calder.. @chrismoya86
2 Followers 432 FollowingEmerson Macedo @emerleite
1K Followers 917 Following Tech Advisor, Problem Solver, VC and AI Specialist. Founder of Be Tech AI and Partner at https://t.co/9A8DcEySer, https://t.co/OxVXJJEBHm and https://t.co/dtgQI61Vtyabderrahim zine @abderrahimzine6
23 Followers 616 FollowingPremraj Thakur @PremrajThakur8
24 Followers 2K Following(((ل()(ل() 'yoav))).. @yoavgo
46K Followers 2K FollowingPercy Liang @percyliang
49K Followers 408 Following Associate Professor in computer science @Stanford @StanfordHAI @StanfordCRFM @StanfordAILab @stanfordnlp | cofounder @togethercompute | PianistKyunghyun Cho @kchonyc
61K Followers 2K Following a combination of a mediocre scientist, a mediocre manager, a mediocre advisor & a mediocre PC at @nyuniversity (@CILVRatNYU) & @genentech (@PrescientDesign).Sam Bowman @sleepinyourhat
35K Followers 3K Following AI alignment + LLMs at NYU & Anthropic. Views not employers'. No relation to @s8mb. I think you should join @givingwhatwecan.Sasha Rush @srush_nlp
52K Followers 464 Following Professor, Programmer in NYC. Cornell Tech, Hugging Face 🤗 https://t.co/cZl0wTfqGzChristopher Manning @chrmanning
126K Followers 115 Following Director, @StanfordAILab. Assoc. Director, @StanfordHAI. Founder, @stanfordnlp. Prof. CS & Linguistics, @Stanford. IP @aixventureshq. 🇦🇺 Do #NLProc & #AI. 👋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/f9QkzO5kaCTal Linzen @tallinzen
16K Followers 893 Following Professor @nyuling and @NYUDataScience, research scientist @GoogleAIGraham Neubig @gneubig
31K Followers 586 Following Associate professor at CMU, studying natural language processing and machine learning.Natasha Jaques @natashajaques
25K Followers 1K Following Senior Research Scientist at @GoogleAI and Assistant Professor @uwcse. Social Reinforcement Learning in multi-agent and human-AI interactions. PhD from @MIT.Felix Hill @FelixHill84
9K Followers 777 Following Research Scientist, Deepmind I try to think hard about everything I tweet, esp on 90s football and 80s music None of my opinions are really someone else'sEdward Grefenstette @egrefen
36K Followers 774 Following FR/US/GB AI/ML Person, Director of Research at @GoogleDeepMind, Honorary Professor at @UCL_DARK, @ELLISforEurope Fellow. All posts are personal.Naomi Saphra @nsaphra
7K Followers 1K Following Waiting on a robot body. ML/NLP. All opinions are universal and held by both employers and family. Same username on every lifeboat off this sinking ship.Zachary Lipton @zacharylipton
59K Followers 2K Following Professor: CMU/@acmi_lab, CTO / CSO: @AbridgeHQ, Creator: @d2l_ai & https://t.co/QQt98VNLUp, Relapsing 🎷Kayo Yin @kayo_yin
8K Followers 556 Following PhD student @berkeley_ai @berkeleynlp working on interpretability and signed languages. Former @msftresearch @deepmind @carnegiemellon @polytechnique. 🇫🇷🇯🇵Christopher Potts @ChrisGPotts
11K Followers 620 Following Stanford Professor of Linguistics and, by courtesy, of Computer Science, and member of @stanfordnlp and @StanfordAILab. He/Him/His.Shane Gu @shaneguML
28K Followers 1K Following Research Scientist & Manager @GoogleDeepMind Tokyo/MTV. ex: @GoogleAI Brain, @OpenAI. (JP: @shanegJP)Ana Marasović @anmarasovic
4K Followers 604 Following Asst prof @UUtah · Ex @allen_ai @uwnlp postdoc @HD_NLP PhD · she/her 🇭🇷Michael Gerovitch @MGerovitch
21 Followers 60 FollowingSonglin Yang @SonglinYang4
2K Followers 2K Following PhD student @MIT_CSAIL. Prev. @ShanghaiTechUni @SUSTechSZ. Working on scalable and principled methods in #ML & #NLProc. INTP | 5w4 | sx/sp | she/herLiam Bright @lastpositivist
64K Followers 5K Following Aspiring philosopher; tolerable human; "amusing combination of sardonic detachment & literally all the feelings felt entirely unironically all at once" [he/his]Chengxu Zhuang @ChengxuZhuang
435 Followers 204 Following Make artificial intelligence models more human like! ICoN Postdoctoral Fellow at MIT. Previously Stanford PhD student @NeuroAILab.Kenneth Li @ke_li_2021
720 Followers 418 FollowingThe TWIML AI Podcast @twimlai
13K Followers 2K Following This Week in #MachineLearning & #AI (podcast) brings you the most interesting and important stories from the world of #ML and artificial intelligence.Swarat Chaudhuri @swarat
2K Followers 543 Following Professor @UTCompSci. Automated Reasoning + Machine Learning + Formal Methods. @iclr_conf Program Chair.Luca Soldaini 🎀 @soldni
6K Followers 1K Following I like tokens! Lead for OLMo data team at @allen_ai (makin Dolma 🍇), open source science fan, @QueerInAI organizer 🤖☕️🍕they/themShakir Mohamed @shakir_za
44K Followers 1K Following ML with Social Purpose. @[email protected] | Research Scientist @DeepMind | Strengthening African ML @DeepIndaba. He/Him. South African 🇿🇦🏳️🌈🌍Edoardo Ponti @PontiEdoardo
2K Followers 389 Following Assistant Professor in #NLP at @EdinburghUni and affiliated lecturer @Cambridge_Uni | Modular deep learningSebastian Gehrmann @sebgehr
5K Followers 2K Following Head of NLP, CTO office, @Bloomberg. (he/him) Generating natural language, one word at a time. Also making sense of that language afterwards. views my ownJascha Sohl-Dickstein @jaschasd
19K Followers 623 Following Member of the technical staff @ Anthropic. Most (in)famous for inventing diffusion models. AI + physics + neuroscience + dynamics.Simran Arora @simran_s_arora
2K Followers 212 Following CS PhD student at @StanfordAILab @hazyresearchYoon Kim @yoonrkim
245 Followers 500 FollowingEve Fleisig @enfleisig
373 Followers 331 Following PhD student @Berkeley_EECS | Princeton ‘21 | NLP, ethical + equitable AI, and linguistics enthusiastRichard Ngo @RichardMCNgo
35K Followers 1K Following What would we need to understand in order to design an amazing future? Figuring that out @openaiKarel D’Oosterlinck @KarelDoostrlnck
2K Followers 593 Following Interpretable AI, RAG, Biomedical NLP. Intern @ContextualAI, PhD student @ugent, visitor @stanfordnlp. Instigator of hikes.Catherine Dulac @DulacLab
9K Followers 985 Following Identifying the neural basis of innate social behaviors using molecular and genetic tools @Harvard @HHMINEWS and at DulacLab at mstdn dot scienceConference on Languag.. @COLM_conf
2K Followers 6 Following https://t.co/GhGCMEoa4A Abstract submission: March 22, 2024Kanishk Gandhi @gandhikanishk
921 Followers 692 Following Phd @Stanford CS; w/ Noah Goodman, Dorsa Sadigh | Prev: @LakeBrenden @NYUDataScience, @IITKanpur, @Path_AIAlex Warstadt @a_stadt
1K Followers 452 Following Postdoc @ ETH Zürich | Future Asst Prof. @ UCSD | Former PhD @ NYU | computational linguistics, NLProc, CogSci, pragmatics | he/him 🏳️🌈Marzieh Fadaee @mziizm
402 Followers 332 Following seeks to understand language. Senior Research Scientist @CohereForAI @Cohere. PhD from @UvA_Amsterdam. [email protected]. Contemplates in private @mzi.Niloofar (Fatemeh) Mi.. @niloofar_mire
4K Followers 1K Following Postdoc @uwcse-@uwnlp, Ph.D. from @ucsd_cse /Privacy, ML, NLP, @winlpworkshop chair, @MSFTResearch - Semantic MachinesAstound Broadband @astoundconnects
26K Followers 6K Following Connecting you to a world of possibilities with our award-winning internet service and 24x7 local customer support. Visit https://t.co/0VHAZh5aZD to get started.Achyuta Rajaram @AchyutaBot
269 Followers 402 Following 17 | mech interp @mit_csail | @atlasfellow '23 | STS 2024Ben Lipkin @ben_lipkin
390 Followers 802 Following phd student @mitbrainandcog. {cogsci, ai} x {language, programs}. he/him.Alex Lew @alexanderklew
644 Followers 839 Following Was a teacher, now a student -- thinking about probability, computation, and pedagogy at MIT's @probcompproj (he/him) 🏳️🌈. (🐘: akl at types dot pl)MIT Graduate Student .. @MITGradUnion
5K Followers 514 Following Working to make MIT a better place for all grad workers. Fellows - sign our Vote Yes petition at https://t.co/3XJKWmG5vh to show MIT admin we need a union for all of us!Arnab Sen Sharma @arnab_api
152 Followers 83 Following Ph.D. student @KhouryCollege, working to make LLMs interpretableNeel 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!Subbarao Kambhampati .. @rao2z
16K Followers 29 Following AI researcher & teacher @SCAI_ASU. Works on Human-Aware AI. Former President of @RealAAAI; Chair of @AAAS Sec T. Here to tweach #AI. YouTube Ch: https://t.co/4beUPOmMW6Ge Yang @EpisodeYang
3K Followers 2K Following I am planting acorns one at a time with policy gradient.thamar | @thamar_solorio
2K Followers 674 Following NLP Prof @MBZUAI, & @UH, Director @RiTUAL_Lab. Friend, mother, partner, loves sunny days and live music. EiC @reviewAcl and ARR board. Views are my own.Ziming Liu @ZimingLiu11
5K Followers 623 Following PhD student@MIT, AI for Physics/Science, Science of Intelligence & Interpretability for ScienceJérémy Scheurer @jeremy_scheurer
384 Followers 301 Following Research Scientist working on AI Alignment @apolloaisafety. Previously: @OpenAI (Evals Contractor), @farairesearch, @ETH_en, @nyuniversitynoahdgoodman @noahdgoodman
2K Followers 109 Following Professor of natural and artificial intelligence @Stanford. Research Scientist at @GoogleDeepMind. (@StanfordNLP @StanfordAILab etc)David Krueger @DavidSKrueger
13K Followers 4K Following Cambridge faculty - AI alignment, deep learning, and existential safety. Formerly Mila, FHI, DeepMind, ElementAI, AISI.Anca Dragan @ancadianadragan
8K Followers 178 Following AI safety & alignment at Google DeepMind • associate professor at UC Berkeley EECS • proud mom of an amazing 2yr oldDeepak Pathak @pathak2206
16K Followers 316 Following I study topics in AI (machine learning, robotics & computer vision).Andreea Bobu @andreea7b
1K Followers 402 Following Research Scientist at BDAII and incoming Assistant Professor @MIT, working at the intersection of machine learning, robotics, and human-robot interaction.Andi Peng @TheAndiPenguin
608 Followers 366 Following PhD student @MIT_CSAIL | formerly @MSFTResearch @Yale @WHOSTP | cats are dope.MAIA (A Multimodal Automated Interpretability Agent) is here! 🧵 📝New paper: arxiv.org/abs/2404.14394 🌐Website: …imodal-interpretability.csail.mit.edu/maia/ Agents like MAIA advance automated interpretation of AI systems from one-shot feature description into an interactive regime where hypotheses…
@srush_nlp @akyurekekin [TLDR] One fully specified behavioral definition (there are other options): Given context w, find largest suffix a such that ab occurs in w. Then output the n-gram distribution p(b | a) fit on w.
@srush_nlp @akyurekekin Let w be the context. A natural solution could be to take the largest a such that there exists b != "" s.t. ab in w (maybe up to some max length for a)
At this point I think I'm just going to use the @lambdaviking bat signal 🔦. Will, have you thought about how either of these definitions formalize? x.com/srush_nlp/stat…
I asked a basic question earlier about what an "Induction Head" was and whether a non-Transformer could have one. The clear answer is no / yes, as Induction Heads means two orthogonal things. lesswrong.com/posts/nJqftaco…
@srush_nlp How to define the induction head behaviorally? It's something like: given `ab...a`, predict `b`. But this definition is underspecified in two ways: 1. b-underspec: a could occur many times with different b's 2. a-underspec: there are different suffix options for a
A Multimodal Automated Interpretability Agent This paper describes MAIA, a Multimodal Automated Interpretability Agent. MAIA is a system that uses neural models to automate neural model understanding tasks like feature interpretation and failure mode discovery. It equips a
Thanks @_akhaliq! MAIA is next in our line of work on Interpretability Agents that interrogate other models, and the functions of their components, using iterative experimentation. Project page: …imodal-interpretability.csail.mit.edu/maia/ w/ @TamarRottShaham @f_x_wang @AchyutaBot @evanqed…
A Multimodal Automated Interpretability Agent This paper describes MAIA, a Multimodal Automated Interpretability Agent. MAIA is a system that uses neural models to automate neural model understanding tasks like feature interpretation and failure mode discovery. It equips a
We should be smarter than just scaling! We should create data-efficient algs. Humans are great at this, algs should learn from humans. This is what I have been working on (t.ly/KB818). This is also what our BabyLM is about (babylm.github.io)!
Zuck on Dwarkesh TLDR: AI winter is here. Zuck is a realist, and believes progress will be incremental from here on. No AGI for you in 2025. 1) Zuck is essentially an real world growth pessimist. He thinks the bottlenecks start appearing soon for energy and they will be take…
I'm thrilled to join Princeton's faculty as an assistant professor in the ECE department starting Fall 2025 🐯 Stay tuned for the launch of my lab. We will develop generally helpful robots that learn and plan 🤖
I am incredibly honored to receive a Glushko Dissertation Prize! A huge thank-you goes to: - My dissertation advisors, @tallinzen and @paul_smolensky, for being incredibly supportive throughout my PhD - (continued in next tweet) 1/2
The Cognitive Science Society is thrilled to announce the winners of the 2024 Glushko Dissertation Prize! 🏆 Let’s meet the brilliant minds behind groundbreaking research in Cognitive Science 🧵👇
I got tenure! It was fitting that I got to celebrate with the lab right after the news. Working together for the last 6.5 years has been a blast.
Learning Transformer Programs (arxiv.org/abs/2306.01128 from Princeton NLP) - This paper is neat. Modify transformer arch to be disentangled (concat not add, -residuals), anneal training to be discrete, convert to python code. Doesn't really scale yet but very fun.
The Special Issue (SI) of @jneurolang (the OA flagship journal of the Neurobiology of Language Society @mitpress) on cognitive computational neuroscience of language🧠🤖 is finally out: shorturl.at/hsDFQ Co-edited with AlessandroLopopolo MilenaRabovsky @roger_p_levy🧵1/n
Special issue @jneurolang on leveraging LLMs to study the mind / brain, including some of our work on trying to understand which aspects of the linguistic stimulus—linguistic structure or meaning— contribute to LLM-brain similarity.
The Special Issue (SI) of @jneurolang (the OA flagship journal of the Neurobiology of Language Society @mitpress) on cognitive computational neuroscience of language🧠🤖 is finally out: shorturl.at/hsDFQ Co-edited with AlessandroLopopolo MilenaRabovsky @roger_p_levy🧵1/n
As many have said already, this is a terrible idea. Invest in the inclusion of the Global South instead, that's what's missing. We don't need to entice adolescents to write papers. #NeurIPS2024
This year, we invite high school students to submit research papers on the topic of machine learning for social impact! See our call for high school research project submissions below. buff.ly/43TiTdD
Thrilled to share a review on THE LANGUAGE NETWORK AS A NATURAL KIND—a culmination of ~20 yrs of thinking about+studying language from linguistic, psycholinguistic, and cog neuro perspectives. @NatRevNeurosci rdcu.be/dEylV With the amazing @neuranna @tamaregev 🥳 🧵1/n
This work really shows how inspiration from how children learn can lead to meaningful progress in language model learning efficiency. We are excited to continue this road to produce more human-like learning algorithms! Thx to my great mentors: @jacobandreas @ev_fedorenko 7/7
Two papers! Can visual grounding help LMs learn more efficiently? 1. We show that algs like CLIP don't learn language better (t.ly/eQHA9) 2. We then propose a new one, LexiContrastive Grounding, which does! (t.ly/KB818) Code: t.ly/C0wu- 🧵
Optimal control via options in language based semi-MDPs
is there a research problem that you consider your Chosen Nemesis? the impossible one you have been slowly toiling on for years in the background? secondary to all the work that actually yields fruit, but where it nonetheless places first in your soul
Our second BabyLM Challenge is here! This year, we feature a vision-language track. You can also bring your own data as long as it has less than 100M words! See babylm.github.io for more info. Let's develop baby-like language learning algs!
👶 BabyLM Challenge is back! Can you improve pretraining with a small data budget? BabyLMs for better LLMs & for understanding how humans learn from 100M words New: How vision affects learning Bring your own data Paper track babylm.github.io 🧵