He He @hhexiy
NLP researcher. Assistant Professor at NYU CS & CDS. hhexiy.github.io Joined December 2016-
Tweets99
-
Followers5K
-
Following351
-
Likes252
Today we're launching PRISM, a new resource to diversify the voices contributing to alignment. We asked 1500 people around the world for their stated preferences over LLM behaviours, then we observed their contextual preferences in 8000 convos with 21 LLMs arxiv.org/abs/2404.16019
We released this new agenda on LLM-safety yesterday. This is VERY comprehensive covering 18 different challenges. My co-authors have posted tweets for each of these challenges. I am going to collect them all here! P.S. this is also now on arxiv: arxiv.org/abs/2404.09932
We released this new agenda on LLM-safety yesterday. This is VERY comprehensive covering 18 different challenges. My co-authors have posted tweets for each of these challenges. I am going to collect them all here! P.S. this is also now on arxiv: arxiv.org/abs/2404.09932
How do you know if a method is better, or just has better hyperparameters? @hhexiy, @kchonyc, and I give a new tool to answer this in our #NAACL2024 paper: "Show Your Work with Confidence" arxiv.org/abs/2311.09480. Use it in your own work with just a "pip install opda"! 🧵 1/8
[1/7] Pre-trained LMs can do in-context learning, but this is unexpected given the distribution shift between pre-training data and ICL prompts. What structures of pre-training data yield ICL? Check out our work “Parallel Structures in Pre-training Data Yield In-Context Learning”
How ICL 𝘦𝘮𝘦𝘳𝘨𝘦𝘴 from unsupervised data? 𝘐𝘵 𝘭𝘦𝘢𝘳𝘯𝘴 𝘧𝘳𝘰𝘮 parallel phrases After deleting parallel parts the ICL ability was reduced by 51% deleting random words - only 2% 🧵 @yanda_chen_ @henryzhao4321 @Zhou_Yu_AI @hhexiy @columbianlp arxiv.org/abs/2402.12530
Super thrilled to share our latest work, AlphaGeometry from @GoogleDeepMind , the first AI system ever approaching the IMO gold medalists in solving Olympiad geometry math problems. Published today at Nature, titled “Solving olympiad geometry without human demonstrations”, our…
Proud of this work. Here's my 22min video explanation of the paper: youtube.com/watch?v=TuZhU1…
Proud of this work. Here's my 22min video explanation of the paper: youtube.com/watch?v=TuZhU1…
#neurips2023 LLM+reasoning 🚨 We stress-test LLM deductive reasoning w OOD examples. Reasoning: We study the OG logical reasoning. OOD: GPT-3.5, PaLM, Llama, FLAN-T5 are tested on 1) proofs w different rules (from in-context ones), 2) deeper, 3) wider, 4) compositional proofs.
#neurips2023 LLM+reasoning 🚨 We stress-test LLM deductive reasoning w OOD examples. Reasoning: We study the OG logical reasoning. OOD: GPT-3.5, PaLM, Llama, FLAN-T5 are tested on 1) proofs w different rules (from in-context ones), 2) deeper, 3) wider, 4) compositional proofs.
Hi 🌎! I've arrived at @NeurIPSConf 🫡 Reach out if you wanna talk all things human feedback + sociotechical alignment. I’m presenting this cute poster, but we’re also building an awesome new human feedback dataset (release in Jan 👀) that I can’t wait to tell everyone about🕺
If you are interested in truthfulness/interpretability of LLMs, chat with @javirandor at #NeurIPS2023 !
If you are interested in truthfulness/interpretability of LLMs, chat with @javirandor at #NeurIPS2023 !
Also, I am 1000% hiring PhD students this round! If you want to work on - open models - collaborative/decentralized training - building models like OSS - coordinating model ecosystems - mitigating risks you should definitely apply! Deadline is Friday 😬 web.cs.toronto.edu/graduate/how-t…
Also, I am 1000% hiring PhD students this round! If you want to work on - open models - collaborative/decentralized training - building models like OSS - coordinating model ecosystems - mitigating risks you should definitely apply! Deadline is Friday 😬 web.cs.toronto.edu/graduate/how-t…
🚨 GPQA (Google-proof QA): very difficult Qs designed for scalable oversight (see 🧵); expert perf >> all GPT-4 baselines (including w/ search) > highly-skilled non-expert perf. I’m excited about this >1-yr-long project coming out. Huge effort in collecting high-quality data!
🚨 GPQA (Google-proof QA): very difficult Qs designed for scalable oversight (see 🧵); expert perf >> all GPT-4 baselines (including w/ search) > highly-skilled non-expert perf. I’m excited about this >1-yr-long project coming out. Huge effort in collecting high-quality data!
📢 I am recruiting Ph.D. students for my new lab at @nyuniversity! Please apply, if you want to work on understanding deep learning and large models, and do a Ph.D. in the most exciting city on earth. Details on my website: izmailovpavel.github.io. Please spread the word!
Attention, prospective MS and PhD applicants! Don’t miss our upcoming virtual fall 2024 program admissions information sessions – MS program: Thurs, 10/19 @ 10am Register at: nyu.zoom.us/webinar/regist… PhD program: Thurs, 10/26 @ 1pm, Register at: nyu.zoom.us/webinar/regist…
The NYU Center for Data Science ML² group, created by @sleepinyourhat and @kchonyc, is doing groundbreaking work at the intersection of machine learning & language. We talked to @hhexiy, @tallinzen, & @JoaoSedoc and heard all about it! #NLP #datascience nyudatascience.medium.com/machine-learni…
🚨AI-assisted writing is a nebulous space with researchers and writers being both curious & apprehensive. How useful are current SOTA LLMs to writers?What are their needs & expectations?If this intrigues you, then buckle up🚨 Paper: arxiv.org/pdf/2309.12570… #NLProc #HCI #ChatGPT
Check out our fascinating interview with CDS PhD student Vishakh Padmakumar @vishakh_pk, who recently organized an #ACL Student Research Workshop and had his first-author paper (with @hhexiy, @ank_parikh, & Richard Yuanzhe Pang) accepted at #ICML. #ai #NLP nyudatascience.medium.com/meet-the-resea…
(((ل()(ل() 'yoav))).. @yoavgo
46K Followers 2K FollowingKyunghyun 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).Percy Liang @percyliang
49K Followers 408 Following Associate Professor in computer science @Stanford @StanfordHAI @StanfordCRFM @StanfordAILab @stanfordnlp | cofounder @togethercompute | PianistSam 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.William Wang @WilliamWangNLP
14K Followers 716 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.Sasha Rush @srush_nlp
52K Followers 464 Following Professor, Programmer in NYC. Cornell Tech, Hugging Face 🤗 https://t.co/cZl0wTfqGzAkari Asai @AkariAsai
11K Followers 650 Following Ph.D. student @uwcse & @uwnlp. NLP. IBM Ph.D. fellow (2022-2023). Meta student researcher (2023-) . ☕️ 🐕 🏃♀️🧗♀️🍳Graham Neubig @gneubig
31K Followers 586 Following Associate professor at CMU, studying natural language processing and machine learning.Jacob Andreas @jacobandreas
14K Followers 958 Following Teaching computers to read. Assoc. prof @MITEECS / @MIT_CSAIL (he/him). https://t.co/5kCnXHjtlY https://t.co/2A3qF5vdJwDelip Rao e/σ @deliprao
46K Followers 5K Following Busy inventing the shipwreck. @Penn. Past: @johnshopkins, @UCSC, @Amazon, @Twitter ||Art: #NLProc, Vision, Speech, #DeepLearning || Life: 道元, improv, running 🌈Tal Linzen @tallinzen
16K Followers 894 Following Professor @nyuling and @NYUDataScience, research scientist @GoogleAINaomi 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.Mark Dredze @mdredze
4K Followers 786 Following John C Malone Professor at @JohnsHopkins @JHUCompSci @jhuclsp @jhumceh; Part time @techatbloomberg (tweets my own) Mastodon @[email protected]Danish Pruthi @danish037
7K Followers 628 Following Faculty at Indian Institute of Science, Bangalore. PhD from @LTIatCMU.Bill Yuchen Lin 🤖 @billyuchenlin
6K Followers 2K Following Research @allen_ai. I evaluate (multi-modal) LLMs, build agents, and study the science of LLMs. Previously: @GoogleAI & @MetaAI FAIR @nlp_uscLuca 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/themWenhu Chen @WenhuChen
11K Followers 520 Following AI researcher @UWaterloo @GoogleAI @VectorInst. Interested in natural language processing, diffusion models. I direct TIGER-Lab at UWaterloo.Nathan Schneider @complingy
4K Followers 1K Following Computational Linguist and Professional Nerd at Georgetown University he/him pronouns, ALL the prepositions @[email protected] @complingy.bsky.socialRosanne Liu @savvyRL
33K Followers 966 Following Cofounded & running @ml_collective. Host of Deep Learning Classics & Trends. Research at Google DeepMind. DEI/DIA Chair of ICLR & NeurIPS. Writing https://t.co/IbycyGfnDRNihal Jain @Nihal_Jain_1
72 Followers 266 Following Applied Scientist, AWS CodeWhisperer. Former ML grad student @CarnegieMellonArhant Chaterjee @ArhantC69420
106 Followers 832 FollowingKB @katiebowles_
641 Followers 5K Following Advancing AI for Healthcare at Scale at @AbridgeHQ | $150M Series C 🚀 | We're Hiring!Spencer Bolaños @SpencerClintonB
7 Followers 380 FollowingDacheng Li @DachengLi177
619 Followers 476 Following Intelligence. PhD @Berkeley_EECS @lmsysorg @ucbrise @berkeley_ai, Prev. @Google @SCSatCMU.Rafael Mosquera @23rd_Conspiracy
68 Followers 429 FollowingNick Mumero @nickdee96
131 Followers 1K Following Cofounder at Continuum Ads. Focusing on NLP, Simulation Modelling and Optimization.HolyDifficult @HolyDifficult
9 Followers 144 FollowingZhaotian Weng @WengZhaotian612
0 Followers 9 FollowingPoshak Pathak @PathakPoshak
20 Followers 548 Following Computer Science and Mathematics at University of Louisiana Monroe. Founder/President of ULM Chess Club.Pensé FFun @inftyCategory
113 Followers 6K FollowingShawn Yuxuan Tong @tongyx361
22 Followers 130 FollowingJunwei @JDI_LINK
400 Followers 5K Following Angel Investor. Ph.D. Computer Vision and Parallel ComputingLiangze Jiang @LiangzeJ
19 Followers 288 Following CS PhD student @EPFL_en 🇨🇭 | Previously student researcher @google , undergrad @UESTC1956 🇨🇳Coby Simmons @cobysim
46 Followers 199 FollowingAlo @Hal90910
0 Followers 2K FollowingBob Zocs @bob_zocs
20 Followers 100 FollowingJordan @Jordan31523422
58 Followers 87 Following believe in God never give up best movie you need to watch 2024Nitarshan Rajkumar @nitarshan
799 Followers 1K Following Adviser to the Secretary of State @scitechgovuk. Co-founder @aisafetyinst. Co-created AI Safety Summit and UK AI Research Resource. PhD @cambridge_clKonrad Seifert @praeterpropter
417 Followers 423 Following Experience maximalist trying to understand and improve. Into international cooperation @longtermgov (anon) feedback appreciated: https://t.co/mv6sPPtQgmclaudia @claudmi
40 Followers 723 FollowingDr. Peter S. Park ⏸.. @dr_park_phd
1K Followers 781 Following AI Existential Safety Postdoctoral Fellow @MIT, @Tegmark Lab. @Harvard PhD '23, @Princeton '17. Alum of @JoHenrich Lab. Studies cognition (both human and AI).AI Safety Events and .. @AISafetyEvents
193 Followers 909 Following Newsletter listing upcoming AI safety events and training programs, weekly. https://t.co/8GbW14fJxWMalvina Nikandrou @MNikandrou
67 Followers 410 Following PhD student @EDINrobotics working on Vision and LanguageKishlay Jha @kishlayjha13
315 Followers 1K FollowingNirupama Ratna @ratna_kandala
182 Followers 1K Following Ph.D. student in Linguistics @ IIT Hyderabad BS-MS in Systems Biology #NLP#AI#NeuroscienceLongxuan Yu @Loy004Yu
4 Followers 17 Following𝕋𝕒𝕥𝕤𝕦�.. @tatsuru_kikuchi
355 Followers 3K Following Research Officer at Faculty of Economics, The University of Tokyo. Keywords: Entrepreneur/OpenAI/Quantum/Crypto/Analytics/Consulting. Views are my own. Shiqi Lou @lou_shiqi60535
12 Followers 119 FollowingXueliang Zhao @xz_hku
1 Followers 103 Followingmichael @_michaelginn
183 Followers 265 Following PhD student at @CuLinguistics and @BoulderNLP. Studying NLP and language technology for endangered, low-resource, and Indigenous languages.Wendi Li @windy_lwd
0 Followers 50 FollowingClayton @cthorrez
1K Followers 1K Following LLM applied scientist by day, esports data scientist for fun. Working on rating systems and benchmarks for esports (and LLMs?) I ❤️ paired comparison datali ii iq j @iq_li80427
55 Followers 311 Following(((ل()(ل() 'yoav))).. @yoavgo
46K Followers 2K FollowingKyunghyun 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).Percy Liang @percyliang
49K Followers 408 Following Associate Professor in computer science @Stanford @StanfordHAI @StanfordCRFM @StanfordAILab @stanfordnlp | cofounder @togethercompute | PianistYoav 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/f9QkzO5kaCSam 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.William Wang @WilliamWangNLP
14K Followers 716 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.Sasha Rush @srush_nlp
52K Followers 464 Following Professor, Programmer in NYC. Cornell Tech, Hugging Face 🤗 https://t.co/cZl0wTfqGzGraham Neubig @gneubig
31K Followers 586 Following Associate professor at CMU, studying natural language processing and machine learning.Jacob Andreas @jacobandreas
14K Followers 958 Following Teaching computers to read. Assoc. prof @MITEECS / @MIT_CSAIL (he/him). https://t.co/5kCnXHjtlY https://t.co/2A3qF5vdJwChristopher Manning @chrmanning
126K Followers 115 Following Director, @StanfordAILab. Assoc. Director, @StanfordHAI. Founder, @stanfordnlp. Prof. CS & Linguistics, @Stanford. IP @aixventureshq. 🇦🇺 Do #NLProc & #AI. 👋Delip Rao e/σ @deliprao
46K Followers 5K Following Busy inventing the shipwreck. @Penn. Past: @johnshopkins, @UCSC, @Amazon, @Twitter ||Art: #NLProc, Vision, Speech, #DeepLearning || Life: 道元, improv, running 🌈Tal Linzen @tallinzen
16K Followers 894 Following Professor @nyuling and @NYUDataScience, research scientist @GoogleAINaomi 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.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.Sebastian 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 ownAllen Institute for A.. @allen_ai
53K Followers 361 Following AI for the Common Good. › Join us: https://t.co/DqTs1G4bGO › Get our newsletter: https://t.co/tvb1VpySfLJia-Bin Huang @jbhuang0604
51K Followers 285 Following Associate Professor @umdcs; Part-time Research Scientist @Meta. I like pixels.Greg Durrett @gregd_nlp
6K Followers 752 Following CS professor at UT Austin. I do NLP most of the time. he/himPeter Hase @peterbhase
2K Followers 690 Following Google PhD Fellow at @uncnlp. Interested in interpretable ML, natural language processing, AI Safety, and Effective Altruism.Nicholas Lourie @NickLourie
121 Followers 178 Following I build things. 🤖 Doing a PhD at @nyuniversity (@CILVRatNYU) on better empirical methods for deep learning and data science. Advised by @kchonyc and @hhexiy.Yao Fu @Francis_YAO_
14K Followers 2K Following PhD @EdinburghNLP on LLMs and Machine Reasoning. Ex. @Columbia @PKU1898 @MITIBMLab @allen_ai AGI has yet to come, so keep runningCarlos E. Perez @IntuitMachine
30K Followers 4K Following Artificial Intuition, Fluency & Empathy, DL Playbook, Patterns for Generative AI, Patterns for Agentic AI https://t.co/fhXw0zjxXpArthur @itsArthurAI
2K Followers 603 Following The AI Performance Company, on a mission to make AI better for everyone and build ML technology to drive responsible business results. Tweets written by humans.trieu @thtrieu_
2K Followers 241 Following thinking about thinking. created alphageometry, darkflow. prev: nyu, google brain/deepmindMehran Kazemi @kazemi_sm
1K Followers 497 Following Senior Research Scientist @GoogleAI. Research areas: machine/deep learning, large language models, artificial general intelligence. Views my own.Lerrel Pinto @LerrelPinto
5K Followers 181 Following Assistant Professor of CS @nyuniversity. I like robots!Amirhossein Kazemneja.. @a_kazemnejad
839 Followers 483 Following Grad student in NLP @Mila_Quebec, @mcgillu, and @rllabmcgill. Working on Transformers and generalizationMichael Bernstein @msbernst
16K Followers 2K Following @Stanford, Associate Professor of Computer Science. I design (better) social tech.The Spectator Index @spectatorindex
2.8M Followers 0 Following News, media and data from around the globe. Covering politics, economics, science, tech and sport.Hannah Rose Kirk @hannahrosekirk
3K Followers 683 Following AI researcher trying to make sense of all things cyberspace 🤖 Uni of Ox PhD (loading…) @oiioxford & @OxfordAI. Prev @turinginst & @Cambridge_Uni. Visitor @ NYUEvan Miller @EvMill
5K Followers 160 Following Statistically inclined software developer, occasional blogger about math + stats stuffWeijie Su @weijie444
4K Followers 444 Following Associate Professor @Wharton & CS Penn. coDir @Penn Research #MachineLearning. PhD @Stanford. #Privacy #DeepLearning #Statistics #GameTheory #Optimization.jietang @jietang
2K Followers 62 Following Professor @ Tsinghua University, Artificial Intelligence, Data Mining, Social Network, Knowledge GraphML Safety Daily @topofmlsafety
2K Followers 2 Following ML safety papers as they are released. Course: https://t.co/l0e0Y2i3AU Newsletter: https://t.co/8Y1kh2D7K6 Main Twitter: https://t.co/AXoYPrylddBoaz Barak @boazbaraktcs
17K Followers 419 Following Computer Scientist. See also https://t.co/EXWR5k634w, https://t.co/SEVX6it6z3 ( @[email protected] , boaz.barak in threads ). Opinions my own.Mrinmaya Sachan @mrinmayasachan
2K Followers 2K Following Assistant Professor of Computer Science at ETH Zurich working in natural language processing (#NLProc), machine learning and education (#edtech).Chenghao Yang @chrome1996
947 Followers 625 Following Ph.D. student @UChicago Ex-SR @google Ex-Scientist @AWS. Ex-RA @jhuCLSP @columbianlp @TsinghuaNLP. Ex-Intern @IBMWatson @AWS. Opinions are my own.Neel 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!Sang Michael Xie @sangmichaelxie
3K Followers 709 Following PhD student @StanfordAILab @StanfordNLP @Stanford advised by Percy Liang and Tengyu Ma. Prev: visiting @GoogleAI Brain, BS, MS Stanford ‘17Tim Dettmers @Tim_Dettmers
29K Followers 819 Following PhD Student at @UW. I blog about deep learning and PhD life at https://t.co/Y78KDJJFE7.Shi Feng @ihsgnef
480 Followers 768 Following NYU Alignment Research Group Incoming Asst. Prof. at GWU (Fall 2024)Surge AI @HelloSurgeAI
4K Followers 146 Following Love language? So do we. Surge AI is the world's most powerful data labeling and RLHF platform, designed from the ground up for stunning AI.Brown NLP @Brown_NLP
3K Followers 144 Following Language Understanding and Representation Lab at Brown University. PI: Ellie Pavlick.lmsys.org @lmsysorg
37K Followers 171 Following Large Model Systems Organization. We created Vicuna and Chatbot Arena! Compare 30+ LLMs (GPT-4/Claude/Llamas) side-by-side at https://t.co/IDFeIDIOtmJane Pan @JanePan_
79 Followers 123 Following CS PhD at @nyuniversity, @NSF GRFP, @Deepmind Fellowship, @SiebelScholars | @Princeton @Princeton_nlp '23 | @Columbia '21.Martin Wattenberg @wattenberg
18K Followers 509 Following Human/AI interaction. Visualization as design, science, art. Professor at Harvard, and part-time at Google's People+AI Research initiative.Dimitris Papailiopoul.. @DimitrisPapail
11K Followers 972 Following prof @ wisconsin; thinking about transformers; learning in context; babas of Inez LilyAjeya Cotra @ajeya_cotra
6K Followers 285 Following AI could get really powerful soon and I worry we're underprepared. Analysis+grantmaking in AI alignment @open_phil (views my own), editor+writer @plannedobs.Pavel Izmailov @Pavel_Izmailov
6K Followers 1K Following Incoming Assistant Professor @nyuniversity 🏙️ Previously @OpenAI #StopWar 🇺🇦Ofir Press @OfirPress
10K Followers 3K Following I build tough benchmarks for LMs and then I get the LMs to solve them. Postdoc @Princeton. PhD from @nlpnoah @UW. Ex-visiting researcher @MetaAI & @MosaicML.Yoav Levine @YoavLevine
306 Followers 82 FollowingJuliana Freire @jfreirenet
1K Followers 258 Following Juliana Freire is a Professor at the Department of Computer Science and Engineering and Data Science at New York University.Together AI @togethercompute
27K Followers 303 Following The future of AI is open-source. Let's build together.Hyung Won Chung @hwchung27
18K Followers 229 Following Research Scientist @OpenAI. Past: @Google Brain / PhD @MITRaphaël Millière @raphaelmilliere
10K Followers 2K Following Philosopher of Artificial Intelligence & Cog Science @Macquarie_Uni Past @Columbia @UniofOxford Also on other platforms Blog: https://t.co/2hJjfSid4ZMina Lee @MinaLee__
3K Followers 452 Following Postdoc at @MSFTResearch | Assistant Professor at @UChicagoCS (2024) | PhD at @Stanford | Language models, AI-assisted writing, Human-AI interaction ✍️Stanislav Fort ✨�.. @stanislavfort
10K Followers 6K Following AI @GoogleDeepMind | Stanford PhD in AI & Cambridge physics | ex-{Anthropic, Stability, Google Brain} | techno-optimism+alignment+progress+growth 🇺🇸🇨🇿James Bradbury @jekbradbury
11K Followers 8K Following Compute at @AnthropicAI! Previously JAX, TPUs, and LLMs at Google, MetaMind/@SFResearch, @Stanford Linguistics, @Caixin.ML-NYC Speaker Series @MLNYCSeries
266 Followers 39 Following The ML-NYC Speaker Series and Happy Hour is a new monthly event for NYC-based machine learning researchers to meet and watch talks from leading researchersTsinghua KEG (THUDM) @thukeg
5K Followers 151 Following #ChatGLM #GLM130B #CodeGeeX #CogVLM #CogView #AMiner The Knowledge Engineering Group (KEG) and THUDM at @Tsinghua_Uni @jietang @ericdongyxStanford HAI @StanfordHAI
86K Followers 558 Following The official account of the @Stanford Institute for Human-Centered AI, advancing AI research, education, policy, and practice to improve the human condition.Najoung Kim 🫠 @najoungkim
2K Followers 493 Following At @BULinguistics and visiting @GoogleAI part-time. 🤖🔠🐱Melanie Mitchell @MelMitchell1
44K Followers 655 Following Professor, Santa Fe Institute. More thoughts at https://t.co/nC43NHRozX.Jan Leike @janleike
44K Followers 322 Following ML Researcher, co-leading Superalignment @OpenAI. Optimizing for a post-AGI future where humanity flourishes.Anthropic @AnthropicAI
261K Followers 26 Following We're an AI safety and research company that builds reliable, interpretable, and steerable AI systems. Talk to our AI assistant Claude at https://t.co/aRbQ97uk4d.Defended my thesis yesterday :) Its been a fantastic ride at @columbianlp and I am grateful to my advisor @SmaraMuresanNLP for believing in my work. Special thanks to @VioletNPeng who introduced me to Creative NLG which made a lot of the work in my thesis possible
I do.
Is he travelling the world? 🌍🌏🌎 Yes! 🔥 Does he show up and engage with the students? 👨🏻🏫👩🏻🎓👨🏽🎓 Also yes! 🔥🔥 Does he still advise researchers and colleagues? Indeed! 🔥🔥🔥
1/ Today in Science, we train a neural net from scratch through the eyes and ears of one child. The model learns to map words to visual referents, showing how grounded language learning from just one child's perspective is possible with today's AI tools. science.org/doi/10.1126/sc…
🥈🏆 Our work on LLM backdoors via RLHF poisoning was awarded the 2nd Prize in the Swiss AI Safety Prize Competition organised by @pourdemain_ch!
🧵 Can data poisoning and RLHF be combined to unlock a universal jailbreak backdoor in LLMs? Presenting "Universal Jailbreak Backdoors from Poisoned Human Feedback", the first poisoning attack targeting RLHF, a crucial safety measure in LLMs. 📖 Paper: arxiv.org/abs/2311.14455
I defended the thesis today! Big thanks to my committee @kchonyc @hhexiy @JoaoSedoc @tallinzen, my amazing advisor @sleepinyourhat, and everyone who attended!
Wanted: LLM that turns language description of problem into convex optimization problem
This paper has now been accepted to @acm_chi #CHI2024. Special thanks to @PhilippeLaban who tirelessly supported this project. As a #NLProc person I have always admired the rigor and high evaluation standards in many HCI papers. Happy to play my part. See y’all in Hawaii 🌺🌊🌴
Can #GPT4 ever write fiction that matches the quality of @NewYorker fiction? Bothered by claims about AI surpassing human creativity🤔? Good news🥁:AI is still 3-10X worse at creativity based on our rubric "Torrance Tests for Creative Writing” #NLProc #HCI arxiv.org/pdf/2309.14556…
🚨New paper!🚨 Self-Rewarding LMs - LM itself provides its own rewards on own generations via LLM-as-a-Judge during Iterative DPO - Reward modeling ability improves during training rather than staying fixed ...opens the door to superhuman feedback? arxiv.org/abs/2401.10020 🧵(1/5)
The Conference on Language Modeling 🦙 (colmweb.org) has the mission of "creating a community of researchers with expertise in different disciplines, focused on understanding, improving, and critiquing the development of LM technology." 🧵 Here are 17 papers from 17…
Excited to announce the Workshop on Reliable and Responsible Foundation Models at @iclr_conf 2024 (hybrid workshop). We welcome submissions! Please consider submitting your work here: iclr-r2fm.github.io (deadline: Fed 3, 2024, AOE) Hope to see you in Vienna or…
@hhexiy Thanks for your fantastic work! It is truly inspiring and provides a lot of insights.🚀
Check out our work to learn about a novel way of measuring the deductive reasoning capabilities of LLMs, as well as their failure modes!
Have LLMs mastered deductive reasoning? Check out PrOntoQA-OOD, a synthetic dataset using a complete set of deduction rules. arxiv.org/abs/2305.15269 Stop by the poster on Wed at 10:45-12:45 and ask Abu Saparov all about reasoning (w or w/o LLMs)! #NeurIPS2023
Check out our work on testing deductive reasoning using an "out-of-demonstration" setup! You should also talk to Abu if you're at NeurIPS!
Have LLMs mastered deductive reasoning? Check out PrOntoQA-OOD, a synthetic dataset using a complete set of deduction rules. arxiv.org/abs/2305.15269 Stop by the poster on Wed at 10:45-12:45 and ask Abu Saparov all about reasoning (w or w/o LLMs)! #NeurIPS2023
At NeurIPS this week. Unfortunately no open-link this time, but I'll have a mentor session Thursday at 10am. Here are some fun things to chat about,
We learned on Thursday we needed to put a presentation together, and Sander did a great job. Sander's an undergrad, and this is Sander's first paper, first conference, and first conference talk. Joint work with @ChengleiSi.
EMNLP 2023 Best Theme Paper Ignore This Title and HackAPrompt: Exposing Systemic Vulnerabilities of LLMs Through a Global Prompt Hacking Competition (Sander Schulhoff, Jeremy Pinto et al.) aclanthology.org/2023.emnlp-mai… #EMNLP2023 #NLProc
🧵 New paper: “Personas as a Way to Model Truthfulness in Language Models” We introduce empirical evidence suggesting LLMs may use “personas” to model truthfulness and improve generalization. arxiv.org/abs/2310.18168
@vishakh_pk et al. show that humans using LLMs write more similar content -> less creativity overall. Pretty interesting WRT social media content generation 🎨: arxiv.org/abs/2309.05196
📢Life update:📢 I moved to Toronto, where I'm now an associate professor at the University of Toronto and an associate research director at the Vector Institute. I wrote a blog post about the long winding path that led me here: colinraffel.com/blog/moving-to…
It's nice to see this argument made formal given confusion around this topic. I think an informal version has been obvious to many of us for a long time: you can't give >0 probability to factual but held-out statements w/o having some probability of "hallucinating".
This new paper with Santosh Vempala gives a simple statistical justification for why and when Language Models *should* hallucinate using standard pretraining, even under ideal in-distribution training conditions. [1/7] arxiv.org/abs/2311.14648