🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

Meta PARTNR is a research framework supporting seamless human-robot collaboration. Building on our research with Habitat, we’re open sourcing a large-scale benchmark, dataset and large planning model that we hope will enable the community to effectively train social robots.

How do we enable agents to perform tasks even when these are underspecified? In this work, led by Ram Ramrakhya, we train VLA agents via RL to decide when to act in the environment or ask clarifying questions, enabling them to handle ambiguous instructions ram81.github.io/projects/ask-t…

I’m very excited to introduce Vy, the AI that sees and acts on your computer! It’s a first glimpse of what we’ve been working on at Vercept! Early computers trapped the world's best experts in low-level tasks–loading code, managing memory, fighting errors. Progress

We are out of stealth! So excited to share what we've been building with the world!🥹

Incredibly happy to announce that we've launched Vy, a first look at AI that lives with you on your machine and can interact with it just like you do! Our mission at Vercept is to make interacting with any application on your computer to accomplish any task effortless and

Just show it once. Vy watches you do a task on your computer, then does it for you. Again and again and again and again… 👀➡️🧠➡️⚡Repeatable workflows, no typing required. Our biggest update yet: Watch and Repeat is here.

Huge update: have Vy watch you do a task once and then let it repeat that task for you over and over and over. Two secret features: 1. Easily compose multiple workflows together "Do @[workflow1] and then @[workflow2]" 2. Let Vy make workflows for you "Save a new workflow

🤯 We cracked RLVR with... Random Rewards?! Training Qwen2.5-Math-7B with our Spurious Rewards improved MATH-500 by: - Random rewards: +21% - Incorrect rewards: +25% - (FYI) Ground-truth rewards: + 28.8% How could this even work⁉️ Here's why: 🧵 Blogpost: tinyurl.com/spurious-rewar…

This reminds me a lot of LiT, including the much faster training, so i think it makes a lot of sense for now. But i also think(hope) it will get superseded in a few years, when there's more and better data for joint training as well as better understanding of mixed architectures

Confused about recent LLM RL results where models improve without any ground-truth signal? We were too. Until we looked at the reported numbers of the Pre-RL models and realized they were serverely underreported across papers. We compiled discrepancies in a blog below🧵👇

Thrilled to be sharing an important milestone towards bringing humanoid robots to the home: 1X AI has trained Redwood, a VLA capable of end-to-end mobile manipulation tasks like retrieving objects for users, opening doors, and navigating around the home. Here’s some of its

Had a great time at this CVPR community-building workshop---lots of fun discussions and some really important insights for early-career researchers. I also gave a talk on "Research as an Infinite Game." Here are the slides: canva.com/design/DAGp0iR…

Our Molmo work won Best Paper Honorable mention at #CVPR2025 ! This large project was one of my best experiences with a fantastic team!

Today we’re launching Background Mode!🪄🥂 No more waiting: Tell Vy what to do, and it’ll quietly handle tasks behind the scenes. You keep working, gaming, or just having fun! 🎮