Robin Jia @robinomial
Assistant Professor @CSatUSC | Previously Visiting Researcher @facebookai | Stanford CS PhD @StanfordNLP robinjia.github.io Los Angeles, CA Joined June 2018-
Tweets170
-
Followers3K
-
Following751
-
Likes1K
While studying memorization in LLMs, we struggled with a basic question: can knowledge really be localized to specific parameters? So, we designed two independent benchmarks to answer this. Amazingly, they agree: pruning-based localization is best, works well, but isn’t perfect.
While studying memorization in LLMs, we struggled with a basic question: can knowledge really be localized to specific parameters? So, we designed two independent benchmarks to answer this. Amazingly, they agree: pruning-based localization is best, works well, but isn’t perfect.
Localization in LLMs is often mentioned. But do localization methods actually localize correctly? In our #NAACL2024 paper, we (w/ @_jessethomason_, @robinomial) develop two benchmarking ways to directly evaluate how well 5 existing methods can localize memorized data in LLMs.
Do multimodal foundation models treat every modality equally? Hint: Humans have picture superiority. How about machines? Introducing IsoBench, a benchmark for multimodal models with isomorphic inputs. 🔗 IsoBench.github.io
Join us for the USC Symposium on Frontiers of Generative AI Models in Science and Society! Featuring special guests @alexvespi @nvchawla @YizhouSun @jmuiuc & @robinomial @yuewang314 @ruishanzliu @USCViterbi 📅 Mar 25 📍MCB 101 🔗RSVP (limited space) tinyurl.com/4zpmfysa
Personally my favorite experiment from this data watermark work. BLOOM-176B memorizes SHA hashes from its training data; we can use similar random character sequences to watermark any document collection!
Personally my favorite experiment from this data watermark work. BLOOM-176B memorizes SHA hashes from its training data; we can use similar random character sequences to watermark any document collection!
Determining whether an LLM has trained on your data isn’t a classification problem, it’s a statistical testing problem. Really proud of this work by @johntzwei and @RyanYixiang on using random watermarks to rigorously detect data usage!
Determining whether an LLM has trained on your data isn’t a classification problem, it’s a statistical testing problem. Really proud of this work by @johntzwei and @RyanYixiang on using random watermarks to rigorously detect data usage!
ACL announcement: "The ACL Executive Committee has voted to significantly change ACL's approach to protecting anonymous peer review. The change is effective immediately. (1/4) #NLPRoc
I’m at #NeurIPS23 and on the job market🎷🧳!! Come and talk about anything LLM reasoning, evaluating communicating agents, human-AI collaboration for new discoveries, coffee and jazz in NOLA☕️
hey @emnlpmeeting #EMNLP2023 i lost my bucket hat 😢 it’s white w yellow bananas 🍌 on it like in my profile, pls DM me if found 🙏🏻
I’ll be presenting SCENE at the poster session this *Saturday* at 11am! Come chat about unanswerable questions, minimal pairs, extrapolation, and the value of synthetic data :)
I’ll be presenting SCENE at the poster session this *Saturday* at 11am! Come chat about unanswerable questions, minimal pairs, extrapolation, and the value of synthetic data :)
Sadly I won't be at #EMNLP2023 in person. However, my amazing advisor @robinomial will be presenting this work at the poster session!
Sadly I won't be at #EMNLP2023 in person. However, my amazing advisor @robinomial will be presenting this work at the poster session!
We're hiring! @CSatUSC is growing fast, with multiple openings for tenure-track and tenured positions. 🔍Security/privacy, AI, machine learning, data science, HCI, but exceptional candidates in all areas considered. 📅Deadline: Jan 5 🔗Details: tinyurl.com/5e99bmb8
🚨Now hiring full-time teaching faculty!🚨 Apply now to join our fast-growing, collaborative department, with a brand new building opening this spring. Find out more about our open faculty positions (teaching, tenured, tenure track): cs.usc.edu/about/open-fac… Please share!
We are at #EMNLP2023 this week! 👋 Explore the latest @CSatUSC @USC_ISI research spanning social bias in name translation, AI tools for journalists, ambiguity in LMs, and much more ⬇️ viterbischool.usc.edu/news/2023/12/u… @USCViterbi @emnlpmeeting @USCResearch #NLP
I just arrived in Singapore for #EMNLP2023! 😜 I'll be presenting our work on modeling and predicting LLM capabilities at @GenBench workshop poster session (10am tomorrow). Also presenting it virtually in gather.town at 4pm on Dec 8. Check out our thread below 👇
I just arrived in Singapore for #EMNLP2023! 😜 I'll be presenting our work on modeling and predicting LLM capabilities at @GenBench workshop poster session (10am tomorrow). Also presenting it virtually in gather.town at 4pm on Dec 8. Check out our thread below 👇
It's a great honor to receive this award! Had so much fun gathering together at SoCalNLP today! #SoCalNLP2023 🎉
It's a great honor to receive this award! Had so much fun gathering together at SoCalNLP today! #SoCalNLP2023 🎉
What can we learn from thousands of LLM experiment records? Can we predict emergent abilities? Does this give any clue on making LLM benchmarking more efficient and principled? Check out our paper on analyzing BIG-bench and designing “small-bench.” arxiv.org/abs/2305.14947 🧵⬇️
ICL reduces the need for labeled *training* data, but what about test data? Fantastic work led by USC undergrad @harveyiyun (applying to PhD programs this year!) shows we can estimate ICL accuracy well (on par with 40 labeled examples) using *unlabeled* data + model uncertainties
ICL reduces the need for labeled *training* data, but what about test data? Fantastic work led by USC undergrad @harveyiyun (applying to PhD programs this year!) shows we can estimate ICL accuracy well (on par with 40 labeled examples) using *unlabeled* data + model uncertainties
How do Transformers really do in-context linear regression? 1 TF layer = 3 steps of a second-order method! Can’t be GD, which converges exponentially more slowly. Meanwhile, LSTMs are more like online GD; they don’t learn second-order optimization (likely due to limited memory).
How do Transformers really do in-context linear regression? 1 TF layer = 3 steps of a second-order method! Can’t be GD, which converges exponentially more slowly. Meanwhile, LSTMs are more like online GD; they don’t learn second-order optimization (likely due to limited memory).
Percy Liang @percyliang
49K Followers 408 Following Associate Professor in computer science @Stanford @StanfordHAI @StanfordCRFM @StanfordAILab @stanfordnlp | cofounder @togethercompute | Pianist(((ل()(ل() 'yoav))).. @yoavgo
46K Followers 2K FollowingWilliam Wang @WilliamWangNLP
14K Followers 717 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.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/2A3qF5vdJwJim 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.Akari 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.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_uscKyunghyun 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.Danish Pruthi @danish037
7K Followers 628 Following Faculty at Indian Institute of Science, Bangalore. PhD from @LTIatCMU.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.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/themGreg Durrett @gregd_nlp
6K Followers 752 Following CS professor at UT Austin. I do NLP most of the time. he/himXin 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/himOfir 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.Wenhu Chen @WenhuChen
11K Followers 520 Following AI researcher @UWaterloo @GoogleAI @VectorInst. Interested in natural language processing, diffusion models. I direct TIGER-Lab at UWaterloo.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.HolyDifficult @HolyDifficult
9 Followers 146 FollowingAaron Mueller @amuuueller
754 Followers 547 Following Postdoc with @boknilev and @davidbau ≡ PhD from @jhuCLSP ≡ Into #NLProc, interpretability, and computational psycholinguistics 💻🧠Phillip Lindsay @EastLAPinche
60 Followers 386 FollowingRon Rocket @RonnyRumble
53 Followers 537 Following 🌍 Energy innovator |🌱Renewable energies | 🤖 AI & blockchain visionary | Shaper of the future #Sustainability #TechTrendsKaiser Sun @KaiserWhoLearns
726 Followers 405 Following Ph.D. student at @jhuclsp , human LM that hallucinates. Formerly @MetaAI, @uwnlp and @AWS they/them🏳️🌈Jinyi Ye @JinyiYe6
62 Followers 232 Following Incoming CS PhD student @USC. Research intern @Microsoft. MS in DS, BA in Comm and Econ. Computational social science and human-centered AI.Nick Mumero @nickdee96
131 Followers 1K Following Cofounder at Continuum Ads. Focusing on NLP, Simulation Modelling and Optimization.Qingyun Wang @eagle_hz
126 Followers 379 FollowingRobert L. Logan IV @rloganiv
232 Followers 506 FollowingLam Tung Vo @nolanvo5894
180 Followers 4K FollowingMark R. Hinkle @mrhinkle
7K Followers 5K Following I help enterprises understand and use artificial intelligence. Leveraging my 25 years of enterprise software experience in emerging technology to drive results.Vikram Pande @VikramPande007
23 Followers 353 FollowingAI & Partners @AI_and_Partners
1K Followers 6K Following The European Union AI Act coming into force in 2024 aims to address problems associated with the use, development and/or deployment of AI systems.JHU CLSP @jhuclsp
5K Followers 662 Following Center for Language and Speech Processing at @JohnsHopkins #NLProc #MachineLearning #AI https://t.co/6IXR5OSiDY @[email protected]Eugene Jang @EugeneOnNLP
53 Followers 237 Following Just a boy that like languages. (Senior Staff Researcher at S2W Inc.) Opinions are my own and not the views of my model(s).Jindong Gu @Jindong73504766
287 Followers 886 Following Senior Research Fellow in University of Oxford @OxfordTVG Faculty Researcher @Google #ResponsibleAI #AISafety #GenAI Homepage: https://t.co/YOSVO3jb6hEdmar Miyake @emiyake
38 Followers 462 FollowingAlo @Hal90910
0 Followers 2K FollowingRun-Ze Fan @Vfrz525_
358 Followers 641 Following Research Assistant@GAIR Lab @sjtu1896. NLP/LLMs/Alignment/Instruction Tuning. Looking for a Ph.D. in the 2025 fall (US)Teemu Summanen @teemusum
194 Followers 3K Following Interested in AI, security, healthcare, and Flutter & Dart.👨🏼💻At X for reading diverse views by professionals and hobbyists.🔬📚🫶kenankortan @kenankorta55495
0 Followers 14 FollowingKnowledgeLM Workshop @lm_knowledge
22 Followers 47 Following Towards Knowledgeable Language Models @ ACL 2024 WorkshopSiddharth Tumre (IITD.. @siddharth_tumre
99 Followers 507 Following Masters in Computer Technology at IITD | NLP | Deep LearningBacklinkGPT @BacklinkGPT
12 Followers 103 Following Automate Your 🔗 Link-Building with AI-Personalized Outreach | AI-Driven Outreach Personalization | One-Click Link Prospecting | Automated Contact DiscoveryHarsh Desai @dreamerharsh
1 Followers 3K FollowingArpandeep Khatua @ArpandeepKhatua
2 Followers 66 FollowingArid Hasan @_aridhasan
13 Followers 60 Following Actively Looking for PhD opportunities| GRA & GTA, University of New Brunswick. RI: LLMs, Disinformation, NLProc, CV, and Information Retrieval.Ibrahim Ahmad @Ibrahim63433664
85 Followers 3K FollowingWendi Li @windy_lwd
0 Followers 50 FollowingAtharva Kulkarni @Atharvak311
133 Followers 609 Following MLT grad @LTIatCMU | Ex-NLP/ML research @Apple, @lcs2iiitdFeigaobox @feigaobox
145 Followers 1K FollowingHarsh Kohli @hkohli14
32 Followers 219 Following PhD student at The Ohio State University Previously - Amazon, Microsoft, GeorgiaTech, BITS Pilani Research interests: Language Models and all things NLPPensé FFun @inftyCategory
108 Followers 6K FollowingTanaz Moradi @TanazMoradi28
6 Followers 137 Following PhD. Candidate in Finance | Passed 3 levels of CFA ProgramPercy Liang @percyliang
49K Followers 408 Following Associate Professor in computer science @Stanford @StanfordHAI @StanfordCRFM @StanfordAILab @stanfordnlp | cofounder @togethercompute | Pianist(((ل()(ل() 'yoav))).. @yoavgo
46K Followers 2K FollowingWilliam Wang @WilliamWangNLP
14K Followers 717 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/cZl0wTfqGzJacob Andreas @jacobandreas
14K Followers 958 Following Teaching computers to read. Assoc. prof @MITEECS / @MIT_CSAIL (he/him). https://t.co/5kCnXHjtlY https://t.co/2A3qF5vdJwJim 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.Akari 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.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/f9QkzO5kaCBill 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_uscKyunghyun 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.AI at Meta @AIatMeta
531K 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.Christopher Manning @chrmanning
126K Followers 115 Following Director, @StanfordAILab. Assoc. Director, @StanfordHAI. Founder, @stanfordnlp. Prof. CS & Linguistics, @Stanford. IP @aixventureshq. 🇦🇺 Do #NLProc & #AI. 👋Danish Pruthi @danish037
7K Followers 628 Following Faculty at Indian Institute of Science, Bangalore. PhD from @LTIatCMU.Yi Tay @YiTayML
29K Followers 97 Following chief scientist / cofounder @RekaAILabs 🫠 past: research scientist @google brain 🤯 currently learning to be a dad 🍼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.Greg Durrett @gregd_nlp
6K Followers 752 Following CS professor at UT Austin. I do NLP most of the time. he/himCapt. Caleb S. Willia.. @CaptCSWilliams
30K Followers 16 Following Trojan War Veteran. Homegrown in the District of Columbia. Former Heis' Man Medal of Honor winner. Leader of the 24th Chicago Regiment.Omar Khattab @lateinteraction
11K Followers 2K Following CS PhD candidate @StanfordNLP. 2022 Apple Scholar in AI/ML. Author of ColBERT (https://t.co/2ZtgXoa1np), DSPy (https://t.co/BH7WmMKDXR), & various retrieval & LM systems.chess24 @chess24com
186K Followers 750 Following TePe Sigeman Chess Tournament starts Saturday! ♟️ Abdusattorov, Ju, Arjun, Keymer, Svidler, Grandelius, Korobov, Maurizzi. 📺 9am ET/15:00 CEST/6.30pm IST.Atharva Kulkarni @Atharvak311
133 Followers 609 Following MLT grad @LTIatCMU | Ex-NLP/ML research @Apple, @lcs2iiitdLianhui Qin @Lianhuiq
4K Followers 397 Following Incoming Assistant Professor at UCSD CSE. Currently postdoc at AI2 Mosaic. NLP, ML, AI. I’m recruiting PhD students.Chenhao Tan @ChenhaoTan
4K Followers 902 Following Assistant professor @UChicagoCS @UChicago. Working on human-centered AI, NLP, CSS at @ChicagoHAI, also part of @ChicagoNLP. DM for Postdoc/PhD opportunities.Shayne Longpre @ShayneRedford
4K Followers 998 Following PhD @MIT. Prev: @Google Brain, @apple ML, @stanfordnlp. 🇨🇦 Interests: AI/ML/NLP, Data-centric AI, transparency & societal impactBhavya Vasudeva @bhavya_vasudeva
38 Followers 231 Following PhD candidate @CSatUSC | @iitroorkee'20 | Interested in theory of deep learning, optimization and robustness/generalization | she/herPeter Hase @peterbhase
2K Followers 690 Following Google PhD Fellow at @uncnlp. Interested in interpretable ML, natural language processing, AI Safety, and Effective Altruism.Yangsibo Huang @YangsiboHuang
1K Followers 726 Following PhD candidate @Princeton. Prev: @GoogleAI @AIatMeta.Luca Luceri @LucaLuceri
349 Followers 511 Following Research Scientist @USC_ISI 🤖 I investigate bad stuff on the Internet 🕵🏼 (Kite-)surf amateur, before breaking my neck 🥲Daniel Firebanks-Quev.. @dafirebanks
58 Followers 367 Following M.S student @USC trying to understand LLMs 👨💻|| Recovering musician 🎸Ghazal Khalighinejad @ghazalkhn
41 Followers 172 Following PhD student @duke_nlp @dukecompsci | interested in Reasoning in Foundation Models, Information Extraction, and AI4Sciencemuhtasham @Muhtasham9
1K Followers 907 Following In my pre-training years "Dude! You should move to SF!" - @dylan522pCA High-Speed Rail �.. @CaHSRA
28K Followers 1K Following 🏗 Building the US’s first 200+mph; 100% renewably powered #HSR 📍 SF↔️Central Valley↔️LA, 500+ miles, under 3⃣ hours ⚡️ & decarbonize transportation #BuildHSRDaniel Litt @littmath
37K Followers 871 Following Assistant professor (of mathematics) at the University of Toronto. Algebraic geometry, number theory, forever distracted and confused, etc. He/him.Anastasios Nikolas An.. @ml_angelopoulos
3K Followers 784 Following @Berkeley_EECS Ph.D. with Mike Jordan/Jitendra Malik. Conformal prediction, distribution-free uncertainty quantification, vision/imaging. Former @stanford_ee.Yanda Chen @yanda_chen_
422 Followers 387 Following 3rd year PhD @ColumbiaCompSci, working on NLP & ML | Student Researcher @GoogleAI | Prev Intern @MSFTResearch, @AmazonScienceKatherine Tian @kattian_
712 Followers 494 Following cs/stat @harvard, working on calibration & factuality of LLMs, prev @GoogleAI tensorflow, golden state @warriors fanIshika Singh @Ishika_S_
452 Followers 715 Following PhD student @CSatUSC | @IITKanpur'21 | Research intern @NVIDIAAI | Interested in #EmbodiedAI and #RobotLearningTianyi(Lorena) Yan @LorenaYannnnn
80 Followers 223 Following CS 4th year undergrad @CSatUSC | NLP Research Assistant w/ @robinomial & @muhao_chen | Previously RA @air_tsinghua | SWE Intern @MicrosoftConference on Languag.. @COLM_conf
2K Followers 6 Following https://t.co/GhGCMEoa4A Abstract submission: March 22, 2024Jacob Morrison @jacobcares
206 Followers 265 Following PYI & Policy @ @allen_ai @ai2_allennlp Political opinions are my own, and do not represent the views of AI2derek guy @dieworkwear
815K Followers 964 Following Menswear writer. Editor at @putthison. Creator of @RLGoesHard. Bylines at The New York Times, The Washington Post, The Financial Times, Esquire, and Mr. PorterYiyuan Li not@NeurIPS @YiyuanLi1
301 Followers 375 FollowingTianyu Gao @gaotianyu1350
3K Followers 686 Following CS PhD student @Princeton @Princeton_nlp working on NLP. Previously: @Tsinghua_Uni @TsinghuaNLPMahdi Soltanolkotabi @mahdisoltanol
293 Followers 317 Following Cycling Academic; work on optimization, probability/statistics, theory of deep learning, AI for science/healthcare; director of center on AIF4S @USCEmilio Ferrara @emilio__ferrara
26K Followers 1K Following Prof of CS @CSatUSC interested in using AI for social good! Part of @USCViterbi @USCAnnenberg @USC_ISI @KeckMedicineUSC also: emilioferrara.bsky.socEric @ericmitchellai
4K Followers 485 Following I like AI & music. Working on making LLMs easier & safer to use. Final year PhD student at Stanford advised by Chelsea Finn & Chris Manning.Cas (Stephen Casper) @StephenLCasper
3K Followers 1K Following #AI safety & responsibility. PhD Candidate @ #MIT_CSAIL.The Data Therapist @yuvalmarton
1K Followers 2K Following Computational Linguist, NLP/NLU/AI Research Scientist, Affil. Assistant Professor, tech mentor, corporate emp. Political in sep acnt. Not my employer’s opinionsIrene Chen @irenetrampoline
8K Followers 814 Following ML for equitable healthcare. Assistant Professor @UCBerkeley and @UCSF. Prev @Harvard, @MIT, @MSFTResearchDavid Bau @davidbau
3K Followers 241 Following Computer Science Professor at Northeastern, Ex-Googler. Believes AI should be transparent. @[email protected] @davidbau.bsky.social https://t.co/wmP5LUZRTwAaron Mueller @amuuueller
754 Followers 547 Following Postdoc with @boknilev and @davidbau ≡ PhD from @jhuCLSP ≡ Into #NLProc, interpretability, and computational psycholinguistics 💻🧠Verna Dankers @vernadankers
527 Followers 201 Following PhD student in #NLProc 🇳🇱 👩💻 🏴@Edin_CDT_NLP. interpretability x figurative language processing x compositional generalisation. she/hermanwhohasitall @manwhohasitall
225K Followers 875 Following Satire account. T-shirts: https://t.co/uk8qKmUSs5 Agent: [email protected]. New book on Unbound.kenzi @kenzianidiot
62K Followers 4K Following follow me for my dads weekly agendas! im also medium funny and have a very cute dog. | marketing at @dropout | she/herHexiang (Frank) Hu @Hexiang_Hu
503 Followers 400 Following Research Scientist @GoogleDeepmind | Vision & Language | GeminiCouldn't have asked for a better birthday gift: just received news that I've been appointed as a Research Assistant Professor in the USC Dept. of Computer Science @CSatUSC, starting July 24! Immensely grateful to my mentors, family, and friends for their support in this journey❤️
Some personal updates: I joined OpenAI a few months ago, working on all things robustness/safety/privacy. Also, we are working to publish more of our safety work. See my first project here below, where we make initial progress on prompt injections and other attacks!
Introducing the Instruction Hierarchy, our latest safety research to advance robustness for prompt injections and other ways of tricking LLMs into executing unsafe actions. More details: arxiv.org/abs/2404.13208
Final moments of the magical yet tragic final round game between Ian Nepomniachtchi and Fabiano Caruana. This game will surely go down as one of the greatest games of chess ever played in the Candidates. #chess #chessbaseindia #candidates2024
@robinomial @delliott We see your point @robinomial! We will take it into account while revising our instructions for the author-reviewer discussion period (ongoing work).
Helping out with the new building 👷♂️👷♀️ With @ebiyik_ @swabhz @ibrahim_sabek @ruishanzliu
I just learned from Llama 3 that I finally proved the Birch and Swinnerton-Dyer Conjecture and started my PhD in mathematics at Harvard! Super excited!! 🚀
Excited to share a preview of Llama3, including the release of an 8B and 70B (82 MMLU, should be the best open weights model!), and preliminary results for a 405B model (still training, but already competitive with GPT4). Lots more still to come... ai.meta.com/blog/meta-llam…
Looking forward to welcome @mohitban47 as a distinguished lecturer to @CSatUSC tomorrow and learn about his latest work on Multimodal LLMs: viterbi.usc.edu/calendar/?even… It's going to be a good day at @nlp_usc
Excited & honored to give the Distinguished Lecture at @CSatUSC today! Looking forward to meeting the awesome @nlp_usc group 🙂 -- thanks @swabhz + everyone for the kind invitation 🙏 PS. Details in the link below, feel free to stop by!
Looking forward to welcome @mohitban47 as a distinguished lecturer to @CSatUSC tomorrow and learn about his latest work on Multimodal LLMs: viterbi.usc.edu/calendar/?even… It's going to be a good day at @nlp_usc
.@ReviewAcl Can't we just do away with this confusing intermediary (3) score for meta-reviews? Instead just have a 4 scale AE rating: 1,2,4,5 What's the diff between "major" (3) and "significant" (2) revisions anyway? It only leads to inconsistency between AEs IMO
@robinomial Excellent point @robinomial! Who would admit to having the time to read a five-part response to a 200 word review?
@delliott I think there is too much anonymity in the system. For example I think I did a better job reviewing when I knew my identity would be revealed even just to peer reviewers.
I've just finished Senior Area Chairing for the ARR February 2024 cycle. It was almost a disaster and I think we should talk about our experiences across the community.
New from @GoogleDeepMind: When can you trust your LLM? We show that LLMs consistently overestimate their own accuracy on some topics (eg nutrition) while underestimating it on others (eg math). Our Few-shot Recalibrator fixes LLM over/under-confidence: arxiv.org/abs/2403.18286 🧵
I had a great time visiting UCLA and USC during the spring break to give talks on "Amazing Multilingual Capabilities and Concerning Cultural Biases in Large Language Models". Thanks to @_jessethomason_ @VioletNPeng @kaiwei_chang @robinomial for hosting!
Localization in LLMs is often mentioned. But do localization methods actually localize correctly? In our #NAACL2024 paper, we (w/ @_jessethomason_, @robinomial) develop two benchmarking ways to directly evaluate how well 5 existing methods can localize memorized data in LLMs.
Got blocked by Kramnik yesterday, so thought I would celebrate it by making this parody video :). It is not perfect, but I spent way too many hours on it! I did entertain myself - so I hope I can also entertain you :).
Can we understand & edit unanticipated mechanisms in LMs? We introduce sparse feature circuits, & use them to explain LM behaviors, discover & fix LM bugs, & build an automated interpretability pipeline! Preprint w/ @can_rager, @ericjmichaud_, @boknilev, @davidbau, @amuuueller