David Dohan @dmdohan
reducing perplexity @openai | past: probabilistic programs, proteins, science & reasoning @ google brain 🧠 ddohan.com Joined August 2011-
Tweets571
-
Followers12K
-
Following2K
-
Likes19K
We achieved gold medal-level performance 🥇on the 2025 International Mathematical Olympiad with a general-purpose reasoning LLM! Our model solved world-class math problems—at the level of top human contestants. A major milestone for AI and mathematics.
1/N I’m excited to share that our latest @OpenAI experimental reasoning LLM has achieved a longstanding grand challenge in AI: gold medal-level performance on the world’s most prestigious math competition—the International Math Olympiad (IMO).
I feel this may be helpful to some of you today:
Fun to watch prediction markets update on the news
OpenAI achieved gold medal on 2025 International Math Olympiad (solving 5 of 6 problems)! Thinks for hours and writes proofs in natural language. We've come a long way from LLMs solving 50% of MATH dataset in 2022 Congrats @alexwei_ on spearheading a major milestone!
1/N I’m excited to share that our latest @OpenAI experimental reasoning LLM has achieved a longstanding grand challenge in AI: gold medal-level performance on the world’s most prestigious math competition—the International Math Olympiad (IMO).
How to code a side project in 2025: 1. May 31 - Write project spec 2. Procrastinate 6 months 3. Dec 31 - ask favorite AI to implement it
Scaling pretraining and scaling thinking are two different dimensions of improvement. They are complementary, not in competition.
This is on the scale of the Apollo Program and Manhattan Project when measured as a fraction of GDP. This kind of investment only happens when the science is carefully vetted and people believe it will succeed and be completely transformative. I agree it’s the right time.
Announcing The Stargate Project The Stargate Project is a new company which intends to invest $500 billion over the next four years building new AI infrastructure for OpenAI in the United States. We will begin deploying $100 billion immediately. This infrastructure will secure
@recurrented x.com/dmdohan/status…
@mollyfmielke There's evidence for it: "In all cases, with exception of S9, they report having owned 1-of-3 toys widely sold by Fisher-Price between 1972 and 1989" Anecdotally, friend traces some # colors to license plate on family car. neurocritic.blogspot.com/2013/01/fisher… study: ncbi.nlm.nih.gov/pmc/articles/P…
@SteveMoraco same no more joking on the internet allowed
🚨SCANDAL 🚨 OpenAI trained on the train set for the Millenium Puzzles
these new captchas are getting way too difficult
o3 has literally made 0% progress on the Millennium eval it’s ai winter now
@cHHillee @polynoamial @tamaybes Gotta look for the NP problems of P vs NP: easy to check, hard to do. Not sure what these look like in math outside formal theorem proving
@polynoamial @tamaybes iiuc one of the constraints with FrontierMath is that the results are easy to check. Unless we do it with formal theorem proving, I’m not sure how to do that for unsolved problems Though maybe one tier should be unsolved hard to check ones too
I have yet to find a well-defined task that cannot be optimized by these models. Eval improvement like ARC AGI showcase this dynamic
So we went from 0 to 87% in 5 years in ARC AGI score. There is no wall it seems. GPT-2 (2019): 0% GPT-3 (2020): 0% GPT-4 (2023): 2% GPT-4o (2024): 5% o1-preview (2024): 21% o1 high (2024): 32% o1 Pro (2024): ~50% o3 tuned low (2024): 76% o3 tuned high (2024): 87%
still a ways to go on FrontierMath!
Lots of folks are posting quotes from Gowers/Tao about the hardest split of FrontierMath, but our 25% score is on the full set (which is also extremely hard, with old sota 2%, but not as hard as those quotes imply).
@_xjdr @btc4me2 tbc it's a joke - literally meant it had been 16 hours since previous post & the o1->o3 jump is 32->87% x.com/dmdohan/status…
At this rate, how long til ARC-AGI is “solved”? For context: - gpt-4o @ 5% - Sonnet3.5 @ 14% - o1-preview @ 18% - o1 @ 32% - best scaffolded solution @ 54%
Caveat on the Tao quote: that refers to the hardest "research" split of the dataset, while the 25% is across the entire dataset. x.com/Jsevillamol/st…
@GarrisonLovely To clear a possible misunderstanding: the quotes refer to questions in the highest tier of difficulty of FrontierMath. Not every question in the benchmark is as difficult as the ones Tao and Gowers reviewed.
FrontierMath details: arxiv.org/html/2411.0487…
imo the improvements on FrontierMath are even more impressive than ARG-AGI. Jump from 2% to 25% Terence Tao said the dataset should "resist AIs for several years at least" and "These are extremely challenging. I think that in the near term basically the only way to solve them, short of having a real domain expert in the area, is by a combination of a semi-expert like a graduate student in a related field, maybe paired with some combination of a modern AI and lots of other algebra packages…”
Well, on FrontierMath 2024-11-26 o3 improves the state of the art from 2% to 25% accuracy. These are absurdly hard strongly held out math questions. And on ARC, the semi-private test set and public validation set scores are 87.5% (private) and 91.5% (public). (7/n)
We are used to the cadence of big model releases: GPT2->3->4 took two years each time We’re in a different world now o1 was announced months ago, now already on next generation Expect faster improvement going forward: o1 is like gpt2 if we could jump to gpt4 ~immediately
Michael Nielsen @michael_nielsen
119K Followers 5K Following Searching for the numinous 🇦🇺 🇨🇦, currently live in 🇺🇸 Research @AsteraInstitute https://t.co/maezekzRUb https://t.co/2dWwZKrvrn
Eric Jang @ericjang11
135K Followers 4K Following
Miles Brundage @Miles_Brundage
73K Followers 13K Following AI policy researcher, @lfschiavo wife guy, fan of cute animals and sci-fi, executive director of AVERI (https://t.co/qq9xcmKQas), Substacker, views my own
typedfemale @typedfemale
45K Followers 556 Following a really exciting new account "advanced pytorch user" - @cHHillee alt: @typedalt
Jason Wei @_jasonwei
109K Followers 707 Following ai researcher @meta superintelligence labs, past: openai, google 🧠
Shane Gu @shaneguML
50K Followers 2K Following Gemini/Omni Thinking, Senior Staff AE @GoogleDeepMind. 🇯🇵-born 🇨🇳🇨🇦. ex: Gemini 1.5-3, GPT-4 @OpenAI, Google Robotics (JP: @shanegJP). Personal opinions.
Jay Hack @mathemagic1an
72K Followers 3K Following Head of AI @clickup. Tweets about AI, computing and their impacts on society. Previously founder @codegen / ML @palantir / startups. Not a pseudonym.
Rosanne Liu @savvyRL
53K Followers 1K Following Mom. Cofounded & running @ml_collective. Co-host of Deep Learning Classics & Trends. Research at Google DeepMind. DEI/DIA Chair of ICLR & NeurIPS.
Harrison Chase @hwchase17
115K Followers 567 Following @LangChain Always hiring: https://t.co/D5Ut3loFO7
Jack Clark @jackclarkSF
136K Followers 5K Following @AnthropicAI, ONEAI OECD, co-chair @indexingai, writer @ https://t.co/3vmtHYkIJ2 Past: @openai, @business @theregister. Neural nets, distributed systems, weird futures
James Bradbury @jekbradbury
17K Followers 9K Following Compute at @AnthropicAI! Previously JAX, TPUs, and LLMs at Google, MetaMind/@SFResearch, @Stanford Linguistics, @Caixin.
near @nearcyan
172K Followers 1K Following allow yourself to introspect and realize what was lost: twitter will never return to what it once was. close your phone; think about how you now spend your life
Stanislas Polu @spolu
25K Followers 647 Following co-founder+engineer(https://t.co/SXBR0l9TrF); alumni(https://t.co/z6zJ8xaKGI, https://t.co/CvVTA1CHAo, https://t.co/WOVEe2aLcK, https://t.co/ui9I4Nj7o1);
kipply @kipperrii
11K Followers 1K Following "exhausting in the way that matters. i'd rather chase kipply than wait for someone slower" - claude
Jeff Dean @JeffDean
446K Followers 6K Following Chief Scientist, Google DeepMind & Google Research. Gemini Lead. Opinions stated here are my own, not those of Google. TensorFlow, MapReduce, Bigtable, ...
Maggie Appleton @Mappletons
41K Followers 1K Following Staff Research Engineer at @GitHubNext. But actually a designer. Writes about design, dev, & anthropology. Also on Bluesky 🦋 @maggieappleton.com
Liam Fedus @LiamFedus
35K Followers 1K Following Building industrial-scale science at @periodiclabs Past: VP of Post-Training @OpenAI; Google Brain
rohan anil @_arohan_
43K Followers 2K Following member of technical staff & co-founder of @coreautoai - and continuing to aspire to understand deep learning.
Iman Ghanizada @ImanGhanizada
1K Followers 736 Following working on something new prev: head of autonomic security @ Google
Surjyendu Ray @surjray
74 Followers 2K Following Data Scientist Lead @LexisNexis Risk, AI enthusiast, astronomy/cosmology nerd, Genomics and computational Biology in a past life! Building towards AGI and ASI!
Adam Mainz @MainzOnX
3K Followers 177 Following PyTorch TPU @google | ex @meta | Obsessed with AI & ML Performance | Building what’s next 🔥 | Views are my own
Arpit Jain @arpitjain11
1K Followers 1K Following CEO @splashlearn. Building fearless learners. #keeplearning
Matt Harloff @Matt_Harloff
108 Followers 3K Following
Mike Saeks @Mikeapedia1
174 Followers 2K Following Field CTO @cursor_ai. Previously incubating new businesses @outshiftbycisco and investing @p72vc & @KKR_Co.
Ali Tamaseb @alitamaseb
5K Followers 3K Following People Capitalist. General Partner @DCVC ($5B Deep Tech VC). Author of Super Founders
AgentsofArchitecture @AgentsofArch
4 Followers 404 Following A regional Victorian based Architecture firm that thinks big to support local communities and business.
Tara Rezaei @tararezaeikh
3K Followers 436 Following co-founder @MirendilAI, mit ‘26, prev openai, prod
Arman Adibi @arman_adibi23
704 Followers 3K Following Assistant Professor, @AUG_Cyber |Postdoc @Princeton | Ph.D. from @Penn, @WarrenCntrPenn | Studying machine learning and optimization.
Hossein Sharifi N. �... @Hossein_SHN
670 Followers 3K Following Researcher @RBCBorealis | Ph.D. in Computer Science @SFU | Developing intelligent machines (I think!)| Ex: @VanProstateCtr, @pmcancercentre, @Novartis
Sang Michael Xie @sangmichaelxie
4K Followers 781 Following self-improving @OpenAI. Prev: PhD @StanfordAILab, @GoogleAI Brain/DeepMind, @Meta GenAI
EM @elonamoskova
1 Followers 558 Following
Jacopo Teneggi @JacopoTeneggi
141 Followers 337 Following Computer Science PhD student @JohnsHopkins Research Scientist Intern @PolymathicAI
turboblitz @turboblitzzz
1K Followers 1K Following —dangerously-skip-permissions cofounder: @sundialmd, @selfxyz, @AtelierMissor_ prev research @ethereumfndn
Paco Belso @pacobelso
6 Followers 64 Following
smsmggx @smsmggx
0 Followers 313 Following
mastermind @ainteligentsia
59 Followers 546 Following deep learning student; AI product @ YouTube / Google
Trata (YC W25) @trytrata
971 Followers 405 Following We host conversations that drive public markets
Tasha @TashaPais
3K Followers 3K Following kernel engineer @MatXComputing | prev rl @softmaxresearch, comp sci @RutgersU @Columbia, robot learning @CAIRLab, co founded @heypocket
Sadie Schnierow @sadie_trata
2 Followers 59 Following
JY Z @JunYuZzzzz
39 Followers 4K Following
Loewen Cavill @loewenkc
2K Followers 993 Following CEO @ Amira & 👼 previously @GoogleX (Wing), @CFS_energy & @MIT
Christian Catalini @ccatalini
23K Followers 5K Following Founder w/roots in academia. Founder @MIT Cryptoeconomics Lab. Past: Co-Founder & Chief Strategy Officer, Lightspark. Co-Creator, Libra. Head Economist, Meta.
朱芳文 @zhufangwen
229 Followers 928 Following 我是朱芳文 皖南东至人 枞阳安凤朱氏廿七世 做过汽车油漆工，广告文案，上市公司媒体主编，科技公司公关负责人 在媒体、新消费、新零售领域三次创业失败 欢迎抱团 一起进步 #AI #OPC
AIcontributors @aicontributors
10 Followers 2K Following
张伟 @zhang_wei00544
0 Followers 412 Following
ghagajjbagaikan @FatimaPerea__
0 Followers 222 Following
Michael Y. Li @michaelyli_
668 Followers 370 Following CS PhD @StanfordAILab @StanfordNLP advised by @noahdgoodman and Emily Fox. Prev: undergrad @princeton
Zeel Patel @ZeelMPatel
456 Followers 645 Following @CitSecurities. Prev: @MSFTResearch, @BroadInstitute, @Harvard
Mohit Reddy @MohitReddy13
6K Followers 833 Following Research @OpenAI. Previously: Reasoning lead (CUA, Grok Code Fast, Grok 3, 4) @xAI. Co-Founder @FennelAI (acq by @Databricks), ex TL Google Brain.
Inference Scaled @inferencescaled
1 Followers 43 Following Learn everything you need to know about OpenAI for Business. Codex, ChatGPT, and more!
Kevin Rose @kevinrose
1.5M Followers 2K Following building at @basic_in (@digg) | Podcasts: The Kevin Rose Show, Random Show w/ @tferriss. | Ex: @google, Board of Directors: @ouraring, @hodinkee
sakura @prunusito
2 Followers 587 Following
Gallil Maimon @GallilMaimon
382 Followers 804 Following Speech LMs, Code LMs; PhD @CseHuji, Ex - Meta (FAIR)
Gabriel Synnaeve @syhw
17K Followers 1K Following Nerd & Dad. RL & CodeGen research since before it was cool.
Matt Van Horn @mvanhorn
36K Followers 5K Following Co-founded June (“self-driving oven,” acquired by @webergrills) & co that became @Lyft. Building again, more soon. OS: @slashlast30days 41.6k★ @ppressdev 4.8k★
Umesh Khanna 🇨🇦... @forwarddeploy
13K Followers 4K Following talent eng @xai | 👼 forward deployed angels | 2x founder | @uwaterloo
reece scoggins @romenotinaday
2 Followers 531 Following
adic @adic_9
2K Followers 3K Following meta-optimizer, consciousness experimentalist / hates on viruses / i am not a doctor and this is not medical advice
Michael Nielsen @michael_nielsen
119K Followers 5K Following Searching for the numinous 🇦🇺 🇨🇦, currently live in 🇺🇸 Research @AsteraInstitute https://t.co/maezekzRUb https://t.co/2dWwZKrvrn
Eric Jang @ericjang11
135K Followers 4K Following
Miles Brundage @Miles_Brundage
73K Followers 13K Following AI policy researcher, @lfschiavo wife guy, fan of cute animals and sci-fi, executive director of AVERI (https://t.co/qq9xcmKQas), Substacker, views my own
Aran Komatsuzaki @arankomatsuzaki
181K Followers 365 Following Sharing AI research. Early work on AI (GPT-J, scaling, MoE). Ex ML PhD (GT) & Google.
typedfemale @typedfemale
45K Followers 556 Following a really exciting new account "advanced pytorch user" - @cHHillee alt: @typedalt
Anthropic @AnthropicAI
1.4M Followers 2 Following We're an AI safety and research company that builds reliable, interpretable, and steerable AI systems. Talk to our AI assistant @claudeai on https://t.co/FhDI3KQh0n.
Wojciech Zaremba @woj_zaremba
154K Followers 219 Following AI resilience at OpenAI Foundation Co-Founder of OpenAI https://t.co/OCQ3mpfyyl
Jason Wei @_jasonwei
109K Followers 707 Following ai researcher @meta superintelligence labs, past: openai, google 🧠
Shane Gu @shaneguML
50K Followers 2K Following Gemini/Omni Thinking, Senior Staff AE @GoogleDeepMind. 🇯🇵-born 🇨🇳🇨🇦. ex: Gemini 1.5-3, GPT-4 @OpenAI, Google Robotics (JP: @shanegJP). Personal opinions.
Jay Hack @mathemagic1an
72K Followers 3K Following Head of AI @clickup. Tweets about AI, computing and their impacts on society. Previously founder @codegen / ML @palantir / startups. Not a pseudonym.
Rosanne Liu @savvyRL
53K Followers 1K Following Mom. Cofounded & running @ml_collective. Co-host of Deep Learning Classics & Trends. Research at Google DeepMind. DEI/DIA Chair of ICLR & NeurIPS.
Harrison Chase @hwchase17
115K Followers 567 Following @LangChain Always hiring: https://t.co/D5Ut3loFO7
Jack Clark @jackclarkSF
136K Followers 5K Following @AnthropicAI, ONEAI OECD, co-chair @indexingai, writer @ https://t.co/3vmtHYkIJ2 Past: @openai, @business @theregister. Neural nets, distributed systems, weird futures
James Bradbury @jekbradbury
17K Followers 9K Following Compute at @AnthropicAI! Previously JAX, TPUs, and LLMs at Google, MetaMind/@SFResearch, @Stanford Linguistics, @Caixin.
near @nearcyan
172K Followers 1K Following allow yourself to introspect and realize what was lost: twitter will never return to what it once was. close your phone; think about how you now spend your life
Stanislas Polu @spolu
25K Followers 647 Following co-founder+engineer(https://t.co/SXBR0l9TrF); alumni(https://t.co/z6zJ8xaKGI, https://t.co/CvVTA1CHAo, https://t.co/WOVEe2aLcK, https://t.co/ui9I4Nj7o1);
Mo Bavarian @mobav0
18K Followers 1K Following Scaling up RL at OpenAI 🍓 Optimization and long context research before. Math PhD, MIT.
kipply @kipperrii
11K Followers 1K Following "exhausting in the way that matters. i'd rather chase kipply than wait for someone slower" - claude
Jonas Adler @JonasAAdler
6K Followers 140 Following Research Scientist, DeepMind. AlphaFold, Gemini.
corsaren @corsaren
7K Followers 927 Following i like smart people who disagree with me • Kantian rule consequentialist, compulsive centrist, Ratatouille slatist, non-stochastic parrot
Jubayer Ibn Hamid @jubayer_hamid
1K Followers 195 Following PhD at Stanford CS, SR at Google. Previously: maths/physics at Stanford.
Elorian AI @ElorianAI
2K Followers 7 Following We're a research and product lab, unlocking the next step in AI reasoning capabilities. Visit us at https://t.co/qvNt7MkmMK
Cristian Ponce @c_m_ponce
1K Followers 259 Following Working on building an automated biology lab that you can send experiments to, from your web browser. CEO @ https://t.co/kRn59DDior.
Tara Rezaei @tararezaeikh
3K Followers 436 Following co-founder @MirendilAI, mit ‘26, prev openai, prod
Harsh Mehta @HarshMeh1a
6K Followers 496 Following Co-Founder & CTO @MirendilAI, Past: AI R&D @AnthropicAI, @GoogleDeepmind, Gemini
Mirendil @MirendilAI
4K Followers 0 Following Democratizing frontier AI R&D to accelerate science and technology.
N8 Programs @N8Programs
10K Followers 238 Following Studying Applied Mathematics and Statistics at @JohnsHopkins. Studying In-Context Learning at The Intelligence Amplification Lab.
Vacha @TVachaW
8K Followers 2K Following Writing about tantra, Buddhism, shamanism & a life lived with curiosity. For deeper dives, subscribe to my free Substack: https://t.co/ifQJTX7MVr
TypeSafe AI @typesafeai
18 Followers 7 Following
John Jumper @JohnJumperSci
31K Followers 0 Following
Belinda @belindmo
2K Followers 1K Following founding @sundialmd, under Long Horizon Research. composable agents, version control, long horizon tasks. prev @stanford @stai_research @google @viva_translate
turboblitz @turboblitzzz
1K Followers 1K Following —dangerously-skip-permissions cofounder: @sundialmd, @selfxyz, @AtelierMissor_ prev research @ethereumfndn
Lila Sciences @LilaSciences
4K Followers 0 Following Building scientific superintelligence to solve humankind's greatest challenges.
Rhythm Garg @rhythmrg
4K Followers 479 Following Co-Founder, CTO @appliedcompute 🚂 prev: research @OpenAI @Stanford
Yuandong Tian @tydsh
44K Followers 939 Following Co-founder of @Recursive_SI. ex-Meta FAIR Director. ex-Google. Reasoning, Optimization and Understanding LLM. Novelist in spare time. PhD in @CMU_Robotics.
Blackstar @blackstar_cmptr
2K Followers 7 Following
Hieu Pham @hyhieu226
45K Followers 24 Following Something new 🇻🇳 | ex: @openai, @xai, @augmentcode, @GoogleBrain, @LTIatCMU, @Stanford, ACM ICPC, IMO🥈 Opinions are my own.
Albert Gu @_albertgu
21K Followers 77 Following assistant prof @mldcmu. chief scientist @cartesia_ai. leading the ssm revolution.
General Intelligence ... @intelligenceco
13K Followers 9 Following The General Intelligence Company Of New York - Run an entire company with agents using Cofounder
Ernest Ryu @ErnestRyu
10K Followers 385 Following Assistant Professor at UCLA, Department of Mathematics. Research Scientist at OpenAI.
George Grigorev @iamgrigorev
4K Followers 1K Following pretraining @poolsideai, x- @togethercompute @snap rare specialty coffee lover
David Shor @davidshor
82K Followers 4K Following Head of Data Science at Blue Rose Research, based in NYC, originally from Miami. I try to elect Democrats. Views are my own. he/him🌹
Sam Rose @samwhoo
20K Followers 1K Following That guy who writes visual essays about software at https://t.co/nLO1QnXOPa. Developer Educator @ngrokHQ.
Eli Tyre @EpistemicHope
3K Followers 228 Following Aiming to navigate through the singularity to a humane world. X is approximately a write-only platform for me: https://t.co/mOW9MmEFoG
Max Niederman @MaxNiederman
235 Followers 209 Following crafting artisanal slop as head of quality @ mechanize
CuspAI @cusp_ai
6K Followers 52 Following Pioneering breakthrough materials to accelerate human progress.
Astrid Wilde 🌞 @astridwilde1
17K Followers 8K Following male ☼ we will end human menial labor this decade ☼
Michael Carbin @mcarbin
3K Followers 458 Following Associate Professor in EECS at @MIT | Co-Founder at @unconvai | Founding Advisor at @mosaicml | Programming Systems | Neural Networks | Unconventiona Computing
Ying Sheng @ying11231
17K Followers 861 Following Cofounder & CEO @radixark @lmsysorg | @sgl_project (https://t.co/6e9BrnaWXK) | Do it anyway | Be the light
General Reasoning @GenReasoning
5K Followers 0 Following A long-horizon reinforcement learning company.
Taylor Sorensen @ma_tay_
3K Followers 620 Following Societal Impacts @AnthropicAI | PhD @uwnlp, prev @humansand @stanfordnlp @byuacme | LLMs + {alignment, pluralism, diversity, democracy, human flourishing}
ben hylak @benhylak
51K Followers 3K Following we are hiring. cto @raindrop_ai, prev @apple design, @spacex engYou might like
