Pleased to share new work with @sflippl.bsky.social @eberleoliver.bsky.social @thomasmcgee.bsky.social & undergrad interns at Institute for Pure and Applied Mathematics, UCLA. Algorithmic Primitives and Compositional Geometry of Reasoning in Language Models www.arxiv.org/pdf/2510.15987 🧵1/n

MSR NYC is hiring senior researchers in AI, both broadly in AI/ML & in specific areas (post-training, test-time scaling, modular transfer learning, science of deep learning). aka.ms/msrnyc-jobs We're reviewing on a rolling basis, interviews in Nov/Dec. Please apply here: tinyurl.com/MSRNYCjob

Figuring out how the brain uses information from visual neurons may require new tools, writes @neurograce.bsky.social. Hear from 10 experts in the field. #neuroskyence www.thetransmitter.org/the-big-pict...

Can Only Meat Machines be Conscious? New paper in Trends in Cognitive Sciences, free download until November 26 with this URL: authors.elsevier.com/a/1luwh4sIRv...

Very excited to share that our work (together with co-first author Shanka Subhra Mondal and @neuroai.bsky.social ) on a brain-inspired architecture for planning with LLMs is now out in Nature Communications! www.nature.com/articles/s41... (thread below)

@science.org just dropped a story covering this preprint! Check it out below, and thanks to @cathleenogrady.bsky.social for the great write-up! www.science.org/content/arti...

Reading papers is a basic skill (and dare I say duty?) of scientists, and I include medical doctors in that group. Keeping abreast of the literature is a foundational part of our professions. There aren’t good shortcuts. In any case, reading papers regularly is fun. arstechnica.com/ai/2025/09/s...

Also to clarify, the cost is $100K per year of the visa, so it is actually $300K for a 3-year visa. As postdocs cost about $100K/year, this fee doubles the cost of postdocs on H1-B visas.

🚨Our preprint is online!🚨 www.biorxiv.org/content/10.1... How do #dopamine neurons perform the key calculations in reinforcement #learning? Read on to find out more! 🧵

New book— just got my copy! I know people often don’t read book chapters but I think that’s a mistake. They are usually much more reflective and wide ranging than journal articles.

I have just dropped a new version of my academicdb project, which helps prolific researchers generate a CV automatically. Now with a web interface! If you are interested, please try it out and let me know what you think - it takes a bit of setup work but then runs easily using Docker.

The myth of optimality in human movement science www.sciencedirect.com/science/arti...

🧠 New preprint: Why do deep neural networks predict brain responses so well? We find a striking dissociation: it’s not shared object recognition. Alignment is driven by sensitivity to texture-like local statistics. 📊 Study: n=57, 624k trials, 5 models doi.org/10.1101/2025...

‪@benhayden.bsky.social‬ @tyrellturing.bsky.social‬ @jmgrohneuro.bsky.social‬ @pessoabrain.bsky.social I see a lot of talk on here about how we should avoid "x does y" talk because the brain is "a dynamic, reverberant, reciprocally interconnected system". But this does not follow. A thread...

This paper finds LLMs' ability to understand that others have different beliefs (Theory of Mind) comes from 0.001% of their parameters. Break those specific weights & the model loses both its ability to track what others know AND language comprehension Interesting implications for models (& minds?)

What are your pros/cons of using AI in science? Feel free to respond wrt use or harm in the design of experiments, coding experiments, coding analysis, brain storming analysis, summarizing literature, synthesis of ideas, modeling, novel model development, mathematical proofs, writing, editing.

An ancient and “nearly naked” black hole that astronomers believe may have been created in the first fraction of a second after the big bang has been spotted by the James Webb space telescope. www.theguardian.com/science/2025...

Today is publication day for this beautiful thing. I hope that it will tell a broad audience what historians of science have now long recognized: that alchemy was not a superstitious aberration but rather, an important phase in the history of ideas and of manufacturing.

“In the late 1960s, Dr. Rossiter was working on her Ph.D. at Yale, when a comment from one of her male professors puzzled her. Who, she had asked, were the women in science? There were none, he said. Another professor mumbled something about Marie Curie being the exception.”🙃

Where do some of Reinforcement Learning's great thinkers stand today? Find out! Keynotes of the RL Conference are online: www.youtube.com/playlist?lis... Wanting vs liking, Agent factories, Theoretical limit of LLMs, Pluralist value, RL teachers, Knowledge flywheels (guess who talked about which!)

This summer I have the great pleasure of working w incredible students! Grads: Thomas McGee (UCLA), Sam Lippl (Columbia), Millie Preece (Cambridge), undergrads: Salma, Kimberly, Ziwen, Pierce, the RIPS program at IPAM (Institute for pure & applied math) UCLA. Excited to share our new research soon📝.

Is the hippocampus best understood in term of discrete subfields or functional gradients? Functional gradients are recapitulated within each hippocampal subfield, supporting a role for both discrete & continuously changing computations. Neat work by Bouffard, @barense.bsky.social, & Moscovitch!

"The aim of science is not to open the door to infinite wisdom but to set a limit to infinite error.” - Bertolt Brecht The Life of Galileo

The rumors are true! #CCN2026 will be held at NYU. @toddgureckis.bsky.social and I will be executive-chairing. Get in touch if you want to be involved!

Are we in the new dark age yet?

This is unfortunate: www.nytimes.com/2025/07/28/u... I wrote to the president of Harvard. I hope other faculty will speak their conscience, even if it means more struggle ahead.

A map of the halting problem, illustrated for root-finding methods: thoughtforms.life/halleys-meth...

We’re excited to announce our next keynote speaker: @neuroai.bsky.social (Microsoft Research) 🧠 Ida studies how humans and AI build world models for memory, exploration, and planning. Her talk will show how game-playing AI reveals insights into human and artificial cognition. Join us at #RLC2025!

1- The neuroscience & psychology chapters of Sutton & Barto 2nd edition. full PDF: incompleteideas.net/book/RLbook2... 2- A brief history of intelligence. Not neuroAI, but popular among grad students at RLC & our MSR interns as a deep conceptual introduction. www.abriefhistoryofintelligence.com/book

Propose some socials for RLC! Research topics, affinity groups, niche interests, whatever comes to mind! rl-conference.cc/call_for_soc...

If you're at ICML, check our work on AlgEval, toward algorithmic understanding of generative AI. I couldn't make it in person but am excited to say @taylorwwebb.bsky.social is there presenting our spotlight paper. P.S. If you see Taylor congratulate him on his professorship! arxiv.org/abs/2507.07544

Super excited to share this one!! Meta-learning sparsity and learning rate gives rise to brain-like gradients of complementary learning systems. So complementary learning systems emerge organically through behavior optimization, and it's not just two of them!!

My "Math, Revealed" series is freely available to anyone -- no paywall! -- in the thread below.

I've been building a model of how intrusive thoughts can warp cognitive maps (world model) in PTSD & derail goal-directed behavior. It combines successor representation & prioritized replay. I started off with a grid world, then embeddings. Today in a story recall task, reading the text was moving:

How does the brain stop thoughts? Find out in my article in @natrevneuro.nature.com with Subbu Subbulakshmi & Maite Crespo-Garcia www.nature.com/articles/s41... that integrates 25 yrs of psychology and neuroscience on this vital [email protected] sky.social #neuroskyence #neuroscience

I've been building a model of how intrusive thoughts can warp cognitive maps (world model) in PTSD & derail goal-directed behavior. It combines successor representation & prioritized replay. I started off with a grid world, then embeddings. Today in a story recall task, reading the text was moving:

NIH funding under the new bill (just passed): "Funding for brain research will be consolidated into one institute with a budget of 2.445B. Currently, brain research is funded by several institutes and adds up to over 5B. So it looks like brain research is getting slashed by over 50%." yes? 🧠📈 🧪

🧠 Paper out! We investigated how hippocampal and cortical ripples support memory during movie watching. We found that: 🎬 Hippocampal ripples mark event boundaries 🧩 Cortical ripples predict later recall Ripples may help transform real-life experiences into lasting memories! rdcu.be/eui9l

NEW: NSF will be kicked out of their building. Announcement will be made tomorrow by HUD Sec. and Governor of VA. HUD will take over the NSF building over the next two years. NSF staffer: "There is no planning for NSF, no identified future location, appropriation for a new building or a move."

Dear #mathsky, if you repost this, it will let people see the article for free. Thanks! www.nytimes.com/interactive/...

I recommend everyone consider reading the United Nation’s charter these days. I wonder why we don’t talk about it more. Imagine the world if we truly resolved ourselves to join efforts toward the aims of the UN charter. www.un.org/en/about-us/...

“WE THE PEOPLES OF THE UNITED NATIONS DETERMINED to save succeeding generations from the scourge of war, which twice in our lifetime has brought untold sorrow to mankind, & to reaffirm faith in fundamental human rights, in the dignity & worth of the human person, in the equal rights of men & women”

Our position paper on algorithmic explanations is out—excited to share it! 🙌 Proud of this collaborative effort toward a scientifically grounded understanding of generative AI. @tuberlin.bsky.social @bifold.berlin @msftresearch.bsky.social @UCSD & @UCLA

Pleased to share our ICML Spotlight with @eberleoliver.bsky.social, Thomas McGee, Hamza Giaffar, @taylorwwebb.bsky.social. Position: We Need An Algorithmic Understanding of Generative AI What algorithms do LLMs actually learn and use to solve problems?🧵1/n openreview.net/forum?id=eax...

New paper: "Large Language Models and Emergence: A Complex Systems Perspective" (D. Krakauer, J. Krakauer, M. Mitchell). We look at claims of "emergent capabilities" & "emergent intelligence" in LLMs from the perspective of what emergence means in complexity science. arxiv.org/pdf/2506.11135

Friends, I'm looking for datasets of PTSD narratives over time. This can be diaries, transcripts of real therapy sessions, etc. I've seen some suggestions online but nothing yet that I can actually download and analyze. Any practical pointers to real data that is downloadable appreciated!

Amazing study: @tamir-eliav.bsky.social from the Ulanovsky lab describes replay events in a large scale environment (200 m tunnel). 🦇 www.cell.com/cell/abstrac...

Sometimes the world leaves you no choice but to write poetry.

My latest Aronov lab paper is now published @Nature! When a chickadee looks at a distant location, the same place cells activate as if it were actually there 👁️ The hippocampus encodes where the bird is looking, AND what it expects to see next -- enabling spatial reasoning from afar bit.ly/3HvWSum

Pleased to say "Space, Time, and Memory", an academic book by Oxford University Press edited by the inimitable Lynn Nadel & Sara Aronowitz is now out. I contributed a chapter, "Memory and Planning in Brains and Machines". You can download the entire book for free: library.oapen.org/bitstream/ha...