This paper will be presented as a **Spotlight Talk** at #NeurIPS2024! 🚀🤖 Congratulations to the student lead authors Gunshi Gupta and Karmesh Yadav! 🎉 Check out the thread🧵below!

Introducing EgoMimic - just wear a pair of Project Aria Project Aria @Meta smart glasses 👓 to scale up your imitation learning datasets! Check out what our robot can do. A thread below👇

I'll be at NeurIPS today (East Exhibit Hall A-C #1904) presenting this work as a spotlight poster. Please drop by if you are interested in talking more about diffusion-based representations and world modeling for control!

Excited to share a sneak peek into what we've been building at Yutori! What you see below is our trained model and internal prototype — multiple agents running in parallel in the background, completing tasks of varying complexity, relevant information and cues to step in being

OpenAI's announcement of Operator on Thursday was a great excuse for us to come out of stealth to show off the AI agents tech we've been building at Yutori. Which means I can now say out loud — we're hiring! Our current top hiring priorities are an awesome founding frontend

The best way for robots to learn human tasks is directly from humans! An algorithm developed by a Georgia Tech research team can train humanoid robots through first-person videos. Their project was supported by Project Aria @Meta. @GTResearchNews Georgia Tech Computing bit.ly/3QoPEcF

Thrilled to have contributed to this work, now published in Nature! Check out Microsoft’s blog for live-action gameplay sequences from Muse: microsoft.com/en-us/research…

A UX challenge we’re tackling Yutori is visualizing multi-agent trajectories. With numerous AI model instances working in parallel, the computer-use livestream popularized by single-agent systems (e.g. Claude, Operator) gets very overwhelming. Our first draft instead

One year ago, Abhishek Das and I left Meta to start Yutori. Ten months ago, Dhruv Batra joined us :) Nine months ago, we crystallized our vision. Two months ago, we released a sneak peak into what we’ve been building. Today, can’t be more excited to fully unveil Yutori’s

I started something new last year with a wonderful group of people. We showed a demo in Jan. Today, we’re telling our story — show before you talk! We are re-imagining how people interact with the web — one of humanity’s greatest inventions and a a mess overdue for an

Excited to try out what Yutori has been cooking in the last few months!

Very excited to release our first reasoning model, Magistral. We released the weights of Magistral Small alongside a paper that presents our approach, online RL infrastructure, and findings.

We're excited to launch Scouts — always-on AI agents that monitor the web for anything you care about.

Introducing the world's best (and open) speech recognition models!

new mistral paper just dropped btw

Big vision‑language models often rubber‑stamp flawed agent work, and a simple 2‑step trick called Self‑Grounded Verification cuts those errors and lifts task success. Right now evaluators lean on rigid scripts or a single model pass, so many bad plans skate by. Agreement bias

Mistral Medium 3.1 just landed on Arena leaderboard—punching way above its weight! 🏆 #1 in English (no Style Control) 🏆 2nd overall (no Style Control) 🏆 Top 3 in Coding & Long Queries 🏆 8th overall Small model. Big impact. Try it now on Le Chat and the API!

Magistral from Mistral AI is a model trained to do reasoning on images. Here is a fun challenge : Can you bet Magistral 1.2 at Geoguessr ? 🌎❓ To test that, I recreated an MVP on Hugging Face 🔗 below