Im pleased to share our new work, “Spatio-temporal organization of network activity patterns in the hippocampus”, now out in Cell Reports! With Demi Brizee & David Dupret, we track how rhythms and spiking behavior map onto hippocampal layers. doi.org/10.1016/j.celr…

🧵 1/8 The Illusion of Thinking: Are reasoning models like o1/o3, DeepSeek-R1, and Claude 3.7 Sonnet really "thinking"? 🤔 Or are they just throwing more compute towards pattern matching? The new Large Reasoning Models (LRMs) show promising gains on math and coding benchmarks,

My postdoc work "Unsupervised pretraining in biological neural networks" is out now, along with more than 400GB (milliions neurons) neural data. nature.com/articles/s4158…. Figshare and Github links can be found in the paper.

1/N How do neural dynamics in motor cortex interact with those in subcortical networks to flexibly control movement? I’m beyond thrilled to share our work on this problem, led by Eric Kirk Eric Kirk with help from Kangjia Cai! biorxiv.org/content/10.110…

🚨 Our latest work shows that SOTA VLMs (o3, o4-mini, Sonnet, Gemini Pro) fail at counting legs due to bias⁉️ See simple cases where VLMs get it wrong, no matter how you prompt them. 🧪 Think your VLM can do better? Try it yourself here: vlmsarebiased.github.io/#example-galle… 1/n #ICML2025

How do transformers carry out recurrent computations while being fundamentally feedforward? Excited to present our work on Constrained Belief Updating at #ICML2025, where we show that attention carries out a spectral algorithm in order to parallelize Bayes updating.

Excited to share a connectomics dataset of hippocampus CA3. Explore the dataset: pyr.ai Paper introducing the dataset and some of our discoveries: biorxiv.org/content/10.110…

1/ Today we announce Pleiades, a series of epigenetic foundation models (90M→7B params) trained on 1.9T tokens of human methylation & genomic data. Pleiades accurately models epigenetics for genomic track prediction, generation & neurodegenerative disease detection from cfDNA,

🚀Introducing Hierarchical Reasoning Model🧠🤖 Inspired by brain's hierarchical processing, HRM delivers unprecedented reasoning power on complex tasks like ARC-AGI and expert-level Sudoku using just 1k examples, no pretraining or CoT! Unlock next AI breakthrough with

Half the world will experience a brain disorder in their lifetime. At Nudge, we're building brain interfaces that are safe, precise, and non-invasive to solve that problem. We've raised a $100M Series A led by Thrive Capital and Greenoaks to go faster. We're hiring. Join us.

We've raised a $100M Series A from Thrive Capital, @greenoaks, and a host of great angels. Everyone’s brain can work better. Engineering that idea into reality is the most impactful advancement we can make to life.

Excited to announce our $100M Series A led by Thrive Capital and Greenoaks. If you’re an exceptional (neuro)scientist and are interested in joining please reach out to me directly. We’re expanding the research team across preclinical and human studies and it’s an amazing

The first time I understood the impact of neuromodulation was while witnessing a deep brain stimulation surgery during my first year of medical school. What stuck with me--more than anything else--was the way the patient's face lit up when her symptoms disappeared. Surgical

📍 Robot localisation is the deca-billion-dollar problem no one talks about. Every mobile robot needs to know its location to within 2-5 cm. Sounds simple? It's not. Current solutions force an impossible choice: cheap but inflexible, or smart but fragile. This fundamental

Can synapses in the brain switch their signs between excitatory and inhibitory during learning🚦? Can they act more like weights in artificial neural networks, able to switch signs based on experience 🔃? Excited to share my thesis work in Bernardo Sabatini lab! 🧵 ⬇️ (1/13)

🚨📊ɴᴇᴡ ᴘᴀᴘᴇʀ!! When we learn a new skill—like tying shoelaces or making pasta—we often start with a series of deliberate steps. But eventually, those steps blend into a smooth, single unit: an action chunk. But why and when do we chunk? 👇 📰authors.elsevier.com/a/1lMMD2Hx2-9B8

New electron microscopy study in Nature Neuroscience finds that layer 5 ET neurons mostly connect to inhibitory cells locally, esp. those suppressing other excitatory cells, suggesting a "winner-take-all" system. nature.com/articles/s4159…

People have been asking what we've been up to at Simplex since our transformer geometry post. Short answer: ☑️We figured out how attention implements Bayesian belief updating ☑️where in-context learning comes from ☑️how neural nets use post-quantum computation (!!) Our July

After many yrs building internally Arcadia Science, we're now able to partner more with others on cool organisms. You won't hurt my feelings if you don't read my post, but i will cut you if you don't watch our first organismal spotlight animation -- just skip to the bottom! 🧵

What’s going on inside large AI models? Astera grantees Adam Shai and Paul Riechers are building a new theory of internal structure to better understand intelligence. We sat down with them to learn more about their work as co-founders of Simplex, a research organization: