Thousand Brains Project

Thousand Brains Project

@thousandbrains.org

Advancing AI & robotics by reverse engineering the neocortex. Leveraging sensorimotor learning, structured reference frames, & cortical modularity. Open-source research backed by Jeff Hawkins & Gates Foundation. Explore thousandbrains.org

397 followers611 following361 posts

Thousand Brains Project
@thousandbrains.org

What’s new with Monty? @ramymounir.com walks through recent changes to Monty’s architecture: from separating feature and location updates in the CMP to simplifying Reference Frames. He also discusses ideas to keep object models efficient as Learning Module hierarchies grow.

Thousand Brains Project
@thousandbrains.org

🇮🇹 Three teams, one week in Rome, and three amazing results. Watch the Rome Focus Week hackathon presentations: an ARC-AGI-3 integration, a brand-new attention system prototype in Monty, and new ways to merge object models or split them into reusable parts.

Thousand Brains Project
@thousandbrains.org

So much can happen in two years: 🧠 Major advances in Monty’s capabilities 🛠️ Improved ecosystem and tooling around Monty 🌎 Real-world experiments with robotics hardware arms and ultrasound 🔬 Theoretical advances and a clearer roadmap toward sensorimotor AI

Thousand Brains Project
@thousandbrains.org

🇮🇹 It’s time for the Rome Monty focus week! The team shares hackathon proposals, from unsupervised learning of unfamiliar objects and connecting Monty to ARC-AGI-3 to using symmetry and attention to improve recognition and exploration. Watch the full video here: youtu.be/9vQmx2C7o3M

Thousand Brains Project
@thousandbrains.org

How does Monty perform in the real world? Zachary Danzig shares his undergraduate thesis, which brings Monty to real robotic hardware. His proof of concept project combined stereo vision, depth sensing and a robotic arm for object recognition. Watch the full video here: youtu.be/3wgHNqkRekM

Thousand Brains Project
@thousandbrains.org

📖 Our paper “Thousand-Brain Systems: Sensorimotor Intelligence for Rapid, Robust Learning and Inference” is now published in Neural Computation! Read more about how Monty works, its capabilities, and advantages over transformers: direct.mit.edu/neco/article...

direct.mit.eduThousand-Brains Systems: Sensorimotor Intelligence for Rapid, Robust Learning and InferenceAbstract. Current AI systems achieve impressive performance on many tasks, yet they lack core attributes of biological intelligence, including rapid, continual learning, representations grounded in se...
Thousand Brains Project@thousandbrains.org

/1 🚨 New Paper Drop 🚨 We’ve released “Thousand-Brains Systems: Sensorimotor Intelligence for Rapid, Robust Learning & Inference” — the first working implementation of a thousand-brains system, code-named Monty. 👇 Dive into the full thread 🧵 arxiv.org/abs/2507.04494