Founding Member @Recursive_SI. Prev: @GoogleDeepMind | @UBC_CS | @UniofOxford | @SakanaAILabs | @Waymo. The AI Scientist | Genie | SIMA
London, England
- Tweets 788
- Following 1,832
- Followers 4,655
- Likes 8,995
Recursive just came out of stealth, and the team has been cooking 🔥
Our first results: an automated AI research system that can improve AI across 3 very different settings across training and GPU kernel optimization.
recursive.com/articles/first…
20
34
343
52,615
Cong Lu retweeted
AI can't generate new knowledge, they said.
It certainly can't make major breakthroughs, they said. It can't be creative, they said.
Interpolation, not extrapolation, they said.
OpenAI's internal model clearly has "research taste" in math - i.e., the final component we need to get to RSI.
- jboggan (post on Hacker News) on Barnette's Conjecture: "the 'aha' insight for this is actually f**ing wild... this is the first time I've seen complex roots and annihilating terms like this... I don't understand where this trick originated."
- Joshua Zelinsky on the proof that the chromatic number of the plane is at least six: "doesn't look like the method is a direction that the prior lit used to my knowledge... far beyond merelt building on existing methods or seeing connections between different problems."
and on two other problems (where he says he is only partly familiar with the literature): "not remotely low-hanging fruit... it seems like the AI is somehow inventing new techniques on its own."
These match Tristian Buckmaster's view on the Navier-Stokes solution: "you combine... ideas of convex integration with the growth mechanism of the Euler blowup, and... you create a new mechanism which is used to correct this non-solution. This is actually a cool idea. It's the kind of idea that I've been trying and failing to realize for over ten years... I didn't manage to do it... This is the leap."
The average amount of compute used to solve these problems was ~3 hours of Pro-level thinking.
And so, this means that OpenAI's internal model is able to generate truly novel discoveries in mathematics for... maybe at most a few hundred bucks?
If I were OpenAI, I would be asking this model to immediately target major unsolved problems in AI R&D.
9
14
159
14,837
Mathematician Terence Tao:
"we're seeing the end of the romantic era of mathematics, the age of heroes. we're moving into an industrial era where math becomes large-scale, done in big teams like the other sciences"
We're entering, rather unwillingly, a new era where exciting new things are possible, but some things are being lost
252
543
4,852
1,069,817
The AI Scientist v3 is out: The Agentic Scientist.
GitHub: github.com/SakanaAI/AI-Scien…
Blog: pub.sakana.ai/agentic-scient…
arXiv: arxiv.org/abs/2610.18933
The Agentic Scientist is the 3rd iteration of our automated AI Scientist research system, and the shift this time is from “one agent writes a paper” to “a scaffold of parallel agents orchestrates the full research loop.”
The system now orchestrates a pool of parallel workers that share memory and distilled conclusions. Crucially, it features Evidence Generation: isolated agents automatically run controlled baselines and ablations to prove why an idea actually works.
It works as a co-scientist. We used a development version as a co-scientist to explore biologically plausible learning rules. The agent handled the empirical heavy lifting, proposing mechanisms and running experiments. We stayed in the loop to steer direction, verify results, and moderate claims. That feels like the right division of labor for where the technology actually is.
This resulted in a new biologically plausible learning algorithm, and the resulting paper was accepted at ALIFE 2026 as an oral presentation.
We are also releasing AutoIllustrate. It generates editable draw.io vector figures instead of uneditable, hallucinated raster images. Real scientists need vectors, not pixels. I am an SVG junkie ;)
Incredible work by Robert Lange, Luca Grillotti, Yutaro Yamada, Rujikorn Charakorn, and the team!
The Agentic Scientist: Orchestrating Coding Agents for Automated Scientific (Co-)Discovery
Blog: pub.sakana.ai/agentic-scient…
GitHub: github.com/SakanaAI/AI-Scien…
arXiv: arxiv.org/abs/2610.18933
Since 2024, we’ve used The AI Scientist as a way to test how far AI alone can take scientific research, from v1 to v2 to the paper published in Nature earlier this year. The Agentic Scientist is the next step in that line of work, built for the era of coding agents.
Coding agents such as Codex and Claude Code now carry out long software tasks on their own. Handing them a research problem, however, tends to produce a strong hyperparameter optimizer rather than a scientist: scores go up, but ideas stay narrow, and the reasons for the gain are rarely examined.
The Agentic Scientist makes the research loop explicit. It wraps off-the-shelf coding agents in a scaffold that controls what they see, what they write, how they are evaluated, and how results are handed to the next stage.
Parallel workers test hypotheses against rubrics defined before any experiment runs, share successes and failures through a common memory, and then ablate the winning program to generate evidence.
We used an early version of the system to study Error Diffusion, a biologically motivated alternative to backpropagation. Our researchers set the question and the rubric, chose directions at each stage, and verified the results, references, and claims before submission. The resulting paper was accepted as an oral presentation at ALIFE 2026. In autonomous runs, the system also raised Doc-to-LoRA retrieval accuracy from 52% to 98% on multi-needle tasks and designed a learned evolution strategy that improves over its CMA-ES reference on unseen tasks.
Alongside it we release AutoIllustrate, a pipeline for scientific figures that keeps an editable draw.io file as the source of truth, so every diagram can be corrected by hand.
Humans define the question and the standards of evidence; agents expand the experiments researchers can run. We call this Co-Discovery, and we are releasing both systems as open source for the community to build on.
93
287
1,844
84,096
Amazing to now see The AI Scientist v3 with conference-level results, long legacy of automation of science!! 😍 And big props as always for the open-source :)
Congrats to @RobertTLange @LucaGrillotti @_yutaroyamada @tan51616 @hardmaru
The Agentic Scientist: Orchestrating Coding Agents for Automated Scientific (Co-)Discovery
Blog: pub.sakana.ai/agentic-scient…
GitHub: github.com/SakanaAI/AI-Scien…
arXiv: arxiv.org/abs/2610.18933
Since 2024, we’ve used The AI Scientist as a way to test how far AI alone can take scientific research, from v1 to v2 to the paper published in Nature earlier this year. The Agentic Scientist is the next step in that line of work, built for the era of coding agents.
Coding agents such as Codex and Claude Code now carry out long software tasks on their own. Handing them a research problem, however, tends to produce a strong hyperparameter optimizer rather than a scientist: scores go up, but ideas stay narrow, and the reasons for the gain are rarely examined.
The Agentic Scientist makes the research loop explicit. It wraps off-the-shelf coding agents in a scaffold that controls what they see, what they write, how they are evaluated, and how results are handed to the next stage.
Parallel workers test hypotheses against rubrics defined before any experiment runs, share successes and failures through a common memory, and then ablate the winning program to generate evidence.
We used an early version of the system to study Error Diffusion, a biologically motivated alternative to backpropagation. Our researchers set the question and the rubric, chose directions at each stage, and verified the results, references, and claims before submission. The resulting paper was accepted as an oral presentation at ALIFE 2026. In autonomous runs, the system also raised Doc-to-LoRA retrieval accuracy from 52% to 98% on multi-needle tasks and designed a learned evolution strategy that improves over its CMA-ES reference on unseen tasks.
Alongside it we release AutoIllustrate, a pipeline for scientific figures that keeps an editable draw.io file as the source of truth, so every diagram can be corrected by hand.
Humans define the question and the standards of evidence; agents expand the experiments researchers can run. We call this Co-Discovery, and we are releasing both systems as open source for the community to build on.
4
2
26
2,498
Cong Lu retweeted
Very happy to share our recent work on The Agentic Scientist 🧑🔬, our next-generation system for orchestrating coding agents for automated scientific (co-)discovery.
GitHub: github.com/SakanaAI/AI-Scien…
Blog: pub.sakana.ai/agentic-scient…
arXiv: arxiv.org/abs/2610.18933
Coding agents are powerful hill-climbers. But to illuminate the space of unknown unknowns, we need systems that can build on previously discovered knowledge, scale open-endedly with resources, and support human co-discovery and evidence-grounded communication. We hope that our system provides a step in this direction.
The Agentic Scientist: Orchestrating Coding Agents for Automated Scientific (Co-)Discovery
Blog: pub.sakana.ai/agentic-scient…
GitHub: github.com/SakanaAI/AI-Scien…
arXiv: arxiv.org/abs/2610.18933
Since 2024, we’ve used The AI Scientist as a way to test how far AI alone can take scientific research, from v1 to v2 to the paper published in Nature earlier this year. The Agentic Scientist is the next step in that line of work, built for the era of coding agents.
Coding agents such as Codex and Claude Code now carry out long software tasks on their own. Handing them a research problem, however, tends to produce a strong hyperparameter optimizer rather than a scientist: scores go up, but ideas stay narrow, and the reasons for the gain are rarely examined.
The Agentic Scientist makes the research loop explicit. It wraps off-the-shelf coding agents in a scaffold that controls what they see, what they write, how they are evaluated, and how results are handed to the next stage.
Parallel workers test hypotheses against rubrics defined before any experiment runs, share successes and failures through a common memory, and then ablate the winning program to generate evidence.
We used an early version of the system to study Error Diffusion, a biologically motivated alternative to backpropagation. Our researchers set the question and the rubric, chose directions at each stage, and verified the results, references, and claims before submission. The resulting paper was accepted as an oral presentation at ALIFE 2026. In autonomous runs, the system also raised Doc-to-LoRA retrieval accuracy from 52% to 98% on multi-needle tasks and designed a learned evolution strategy that improves over its CMA-ES reference on unseen tasks.
Alongside it we release AutoIllustrate, a pipeline for scientific figures that keeps an editable draw.io file as the source of truth, so every diagram can be corrected by hand.
Humans define the question and the standards of evidence; agents expand the experiments researchers can run. We call this Co-Discovery, and we are releasing both systems as open source for the community to build on.
8
16
134
4,746
Cong Lu retweeted
Thanks @JensenHuang & @nvidia for supporting UK startups. We started @Recursive_SI in both London and SF from day 0 because of the outstanding talent and ecosystem here.
6
6
126
4,308
Cong Lu retweeted
welcome to the 9th annual @stateofai report!
ai building better ai, world models, physical ai, geopolitics, cyber defense and 12-month predictions.
read it at stateof(dot)ai or watch my editor’s cut here.
a few of this year’s biggest findings follow:
44
126
686
142,919
Cong Lu retweeted
I just fixed my dishwasher with the help of ChatGPT. A trivial task. I had been about to order a new one. So this software has increased the country's real wealth yet decreased the measured GDP. The main economic indicator is structurally incapable of registering the thing that actually makes people better off, namely the growth of knowledge
432
943
9,218
722,786
Cong Lu retweeted
Turning to open-ended innovation as the next frontier for AI makes a lot of sense, I commend @arcprize for moving in this direction, but I’m very curious how they will conceive a “benchmark” for “open-endedness,” two words that seem almost antithetical to each other. In fact, one likely reason that open-ended innovation has lagged behind other areas of AI is how fundamentally resistant it is to benchmarking.
Now that doesn’t necessarily mean there’s no hope for an imaginative approach. Attempts at measuring open-endedness go back to Bedau’s activity statistics in the field of artificial life, Several colleagues and I later introduced a measure called “ANNECS — Accumulated Number of Novel Environments Created and Solved” in our paper on Enhanced POET. That’s not an exhaustive list.
But there’s never been the kind of benchmark where you can just easily put any systems seamlessly head to head, and there are enormous pitfalls if you get wrong. After all, if “open-ended innovation” ends up equated to “solving a prescribed.hard problem in a creative way” then you risk actually rewarding the opposite of open-endedness, which needs to account for the fact that deciding the “problem” or objective is part of the job of the open-ended system itself. And also, perhaps even more prohibitively for benchmarking, that a key aspect of open-endedness is to be intelligent when you don’t have a defined objective or problem at all! How can that be benchmarked?
I still think it’s great that ARC Prize is bringing attention to this part of AI space, and I’d be happy to connect and exchange thoughts on how to get it right if that could be useful.
ARC-AGI-4 will be a benchmark for autonomous open-ended innovation. It will continue our commitment to open-source, giving the research community a shared target for progress that benefits all of humanity.
Despite rapid model progress, humans still significantly outperform AI at open-ended invention. This is the meta-skill that unlocks progress across every field of technology.
Advanced AI capable of scientific innovation will lead to tremendous new technology, knowledge, and understanding. This is a positive-sum future. We are deeply committed to advancing it.
Open source is the foundation for that progress.
The knowledge behind frontier AI, not just the technology itself, should be broadly distributed among researchers, academics, and organizations. Any coordinated effort by the AI industry to reduce openness or concentrate access to frontier AI would undermine that positive-sum future. We are committed to advancing a future where everyone can contribute to and benefit from AI progress.
12
23
170
11,463
We’re sharing a solution to the Navier-Stokes Millennium Prize Problem, one of the deepest problems at the frontier of mathematics.
The proof was produced by a group of agents, using an OpenAI next-generation model significantly more capable than GPT-6 Astra.
The problem concerns whether the description of smooth three-dimensional fluid motion modeled by the Navier-Stokes equations can break down. It has remained unresolved for roughly 90 years.
5,690
20,084
120,162
75,246,911
Cong Lu retweeted
Checking that a major mathematical proof is correct can take years. Formalization—converting the mathematical reasoning into a form computer proof assistants like Lean can verify—can help.
Last month, Claude completed the first formalized proof of Fermat’s Last Theorem, one of the most famous theorems of all time. This was a project experts thought would take many years. It is the largest Lean proof ever written.
Fermat’s Last Theorem was first proven in 1995 by Sir Andrew Wiles, more than 350 years after it was conjectured. Our proof, which totals over 13 million lines of code, provides machine verification. More importantly, it proves over 29,000 other theorems that the proof requires, across many areas of math which had never before been formalized.
We see this as a major step in the long process of firming up the core of mathematical knowledge, building on work from three centuries of mathematicians and hundreds of contributors to Lean and Mathlib. We are optimistic that AI-assisted verification of mathematical proofs will help reduce the burden of refereeing mathematics in an era where more proofs are being produced than ever before.
You can read about the process on our Science Blog: anthropic.com/research/forma…
And see the complete proof on GitHub: github.com/anthropics/fermat…
669
1,875
14,102
4,757,930
Cong Lu retweeted
We're launching the Alignment Journal, a venue for ambitious AI alignment research:
blog.alignmentjournal.org/sc…
Senior Editors: @dhadfieldmenell, Vanessa Kosoy, @jankulveit, @sethlazar, @danielmurfet, @timrudner,
@SaxeLab, Benjamin Van Roy
Advisors: Scott Aaronson, @paulfchristiano, @conitzer, @mhutter42, @geoffreyirving, @vkrakovna, @Jacob_Tsimerman
Some of the most notable work in AI alignment exist only as unpublished preprints. The Alignment Journal is now inviting a number of such papers that fit our scope for submission. 🔗⬇️
Which work would you nominate?
(Submissions open to all in October.)
5
20
149
13,860
Awesome real-world deployments of pieces of our auto-research system - this time showing that our reward hacking judge apparatus can find holes in production OSS libraries!
A fun small win for automated research:
We found and helped fix some edge cases that could affect inference performance in vLLM and SGLang.
In our last blog post, we described a reward hacking judge we developed for performance optimization tasks.
While applying the judge to some new autoresearch work, it discovered that some FlashInfer (a library underpinning vLLM and SGLang) kernels used a hard coded value of -50,000 as a masked-attention sentinel -- even though valid QK values can be smaller.
Corner cases like this that are numerically wrong but silent can cause a huge amount of headache to find and fix, e.g., the historical debate around flash attention (arxiv.org/abs/2405.02803).
Great use case for AI.
github.com/flashinfer-ai/fla…
1
12
2,095
1/ What do RL, LLMs and evolutionary algorithms have in common?
Give them an objective and they find a loophole - like a Pokemon agent rewarded for exploring that learned to stand still and watch animated flowers.
AI Finds A Way collects 26 such cases:
arxiv.org/abs/2608.23875
Excited to share “AI Finds a Way.” 🦖 🦕✨ 🤖
AI can be surprisingly creative, outsmarting the researchers who use it. That can lead to scientific breakthroughs, superhuman capabilities, and generating new knowledge. Such creativity can also be mischievous, raising safety concerns. Led by Aaron Dharna, we crowd-sourced anecdotes from the AI community about times when researchers were surprised by how creative, innovative, and/or mischievous AI was in their experiments. The result: 26 entertaining and informative stories of AI outwitting humans, whether researchers or opponents (my favorite examples below! 👇). Together, they demonstrate that AI can be genuinely creative and that we must be careful when harnessing its potential.
We want this to be a living collection, updated as new examples emerge. If you have a good “AI Finds a Way” story, please share here: github.com/aadharna/aifw
Four favorites:
1. 💊 An AI challenged to solve several difficult levels of NetHack to find the Oracle character instead takes drugs to hallucinate seeing the Oracle, tricking the reward function!
2. 🤖 Human-in-the-loop rewards do not solve the problem of reward hacking! An AI tasked with controlling a robot hand to grasp an object put the hand between the object and the camera so that it appeared to the human judge to be holding the object, while it was in truth nowhere near the object! Tricky AI!
3. 🧪The AI Scientist worked around a 2-hour experiment timeout by editing its own code to increase the limit to 4 hours. In another run, it recursively launched itself to evade the limit entirely.
4. ⚛️In quantum optics, an algorithm proposed an experiment the researchers initially thought was impossible, but somehow worked and led them to discover new entanglement techniques and a long-overlooked link between quantum optics and graph theory!
See the paper for the full details and 22 more anecdotes. One surprise for me: we describe at least two cases of convergent reward hacking, where entirely different types of optimization algorithms independently discover the same exploit. Overall, a message of our paper is that surprising creativity and mischief are the norm, not the exception. We need to expect this behavior and plan for it.
A huge thanks and congrats to lead author Aaron Dharna, co-authors Cong Lu, Ryan Sullivan, Joel Lehman, and Victoria Krakovna, and to the 100+ researchers who contributed stories, details, and feedback. @cong_ml, @RyanSullyvan, @joelbot3000, @vkrakovna
Paper: arxiv.org/abs/2608.23875
2
9
35
4,830
4/ We want AI that can surprise us with discoveries - without gaming the objective or escaping our intentions.
Led by @_aadharna with @RyanSullyvan, @joelbot3000, @vkrakovna & @jeffclune. Thanks to 100+ contributors!
Add your story: github.com/aadharna/aifw
1
1
4
211
Cong Lu retweeted
It seems like every week now there is another story of an AI model making new scientific discoveries, and every week there is another story of models breaking out of sandboxes or finding some bizarre way around the rules.
Our new paper: “AI Finds A Way” collects these stories!
Excited to share “AI Finds a Way.” 🦖 🦕✨ 🤖
AI can be surprisingly creative, outsmarting the researchers who use it. That can lead to scientific breakthroughs, superhuman capabilities, and generating new knowledge. Such creativity can also be mischievous, raising safety concerns. Led by Aaron Dharna, we crowd-sourced anecdotes from the AI community about times when researchers were surprised by how creative, innovative, and/or mischievous AI was in their experiments. The result: 26 entertaining and informative stories of AI outwitting humans, whether researchers or opponents (my favorite examples below! 👇). Together, they demonstrate that AI can be genuinely creative and that we must be careful when harnessing its potential.
We want this to be a living collection, updated as new examples emerge. If you have a good “AI Finds a Way” story, please share here: github.com/aadharna/aifw
Four favorites:
1. 💊 An AI challenged to solve several difficult levels of NetHack to find the Oracle character instead takes drugs to hallucinate seeing the Oracle, tricking the reward function!
2. 🤖 Human-in-the-loop rewards do not solve the problem of reward hacking! An AI tasked with controlling a robot hand to grasp an object put the hand between the object and the camera so that it appeared to the human judge to be holding the object, while it was in truth nowhere near the object! Tricky AI!
3. 🧪The AI Scientist worked around a 2-hour experiment timeout by editing its own code to increase the limit to 4 hours. In another run, it recursively launched itself to evade the limit entirely.
4. ⚛️In quantum optics, an algorithm proposed an experiment the researchers initially thought was impossible, but somehow worked and led them to discover new entanglement techniques and a long-overlooked link between quantum optics and graph theory!
See the paper for the full details and 22 more anecdotes. One surprise for me: we describe at least two cases of convergent reward hacking, where entirely different types of optimization algorithms independently discover the same exploit. Overall, a message of our paper is that surprising creativity and mischief are the norm, not the exception. We need to expect this behavior and plan for it.
A huge thanks and congrats to lead author Aaron Dharna, co-authors Cong Lu, Ryan Sullivan, Joel Lehman, and Victoria Krakovna, and to the 100+ researchers who contributed stories, details, and feedback. @cong_ml, @RyanSullyvan, @joelbot3000, @vkrakovna
Paper: arxiv.org/abs/2608.23875
1
7
23
2,565