Min Liu @minliu01

Self-Questioning Language Models: LLMs that learn to generate their own questions and answers via asymmetric self-play RL. There is no external training data – the only input is a single prompt specifying the topic.

Truly impressive to see a humanoid robot climbing stairs with this level of robustness and reliability!

Skild AI @SkildAI

a month ago

Truly impressive to see a humanoid robot climbing stairs with this level of robustness and reliability!

41 144 1K 247K 145

Download Video

I joined Skild AI late last year and we've been making swift progress towards more general robots! To record these videos we took the robot around town to locations it had never seen before with no prior preparation or planning.

Skild AI @SkildAI

a month ago

I joined Skild AI late last year and we've been making swift progress towards more general robots! To record these videos we took the robot around town to locations it had never seen before with no prior preparation or planning.

41 144 1K 247K 145

Download Video

Modern AI is confined to the digital world. At Skild AI, we are building towards AGI for the real world, unconstrained by robot type or task — a single, omni-bodied brain. Today, we are sharing our journey, starting with early milestones, with more to come in the weeks ahead.…

We’ve been building quietly — starting tomorrow, we go live. Here’s a teaser of what we did before Skild AI. It has shaped what’s coming next. 07/29. Stay tuned.

🚨 The era of infinite internet data is ending, So we ask: 👉 What’s the right generative modelling objective when data—not compute—is the bottleneck? TL;DR: ▶️Compute-constrained? Train Autoregressive models ▶️Data-constrained? Train Diffusion models Get ready for 🤿 1/n

🧠With the shift in humanoid control from pure RL to learning from demonstrations, we take a step back to unpack the landscape. 🔗breadli428.github.io/post/lfd/ 🚀Excited to share our blog post on Feature-based vs. GAN-based Learning from Demonstrations—when to use which, and why it…

Everyone knows action chunking is great for imitation learning. It turns out that we can extend its success to RL to better leverage prior data for improved exploration and online sample efficiency! colinqiyangli.github.io/qc/ The recipe to achieve this is incredibly simple. 🧵 1/N

How can we unlock generalized reasoning? ⚡️Introducing Energy-Based Transformers (EBTs), an approach that out-scales (feed-forward) transformers and unlocks generalized reasoning/thinking on any modality/problem without rewards. TLDR: - EBTs are the first model to outscale the…

TRI's latest Large Behavior Model (LBM) paper landed on arxiv last night! Check out our project website: toyotaresearchinstitute.github.io/lbm1/ One of our main goals for this paper was to put out a very careful and thorough study on the topic to help people understand the state of the…

Your bimanual manipulators might need a Robot Neck 🤖🦒 Introducing Vision in Action: Learning Active Perception from Human Demonstrations ViA learns task-specific, active perceptual strategies—such as searching, tracking, and focusing—directly from human demos, enabling robust…

(1/n) Since its publication in 2017, PPO has essentially become synonymous with RL. Today, we are excited to provide you with a better alternative - EPO.

Training robots for the open world needs diverse data But collecting robot demos in the wild is hard! Presenting DexWild 🙌🏕️ Human data collection system that works in diverse environments, without robots 💪🦾 Human + Robot Cotraining pipeline that unlocks generalization 🧵👇

Low-cost teleop systems have democratized robot data collection, but they lack any force feedback, making it challenging to teleoperate contact-rich tasks. Many robot arms provide force information — a critical yet underutilized modality in robot learning. We introduce: 1. 🦾A…

Excited to share that ManipGen has been accepted to ICRA 2025! We have released the code here: github.com/mihdalal/manip…. Run our generalist system on your robot today!

Murtaza Dalal @mihdalal

10 months ago

Excited to share that ManipGen has been accepted to ICRA 2025! We have released the code here: github.com/mihdalal/manip…. Run our generalist system on your robot today!

10 118 622 145K 359

Download Video

We introduce Dexterity Gen (DexGen), a foundation controller that enables unprecedented dexterous manipulation capabilities. For the first time, it allows human teleoperation of tasks such as using a pen, screwdriver, and syringe. Developed by @berkeley_ai and @metaai. A Thread.

Teaching bimanual robot hands to perform very complex tasks has been notoriously challenging. In our work, Bidex: Bimanual Dexterity for Complex Tasks, we’ve developed a low-cost system that completes a wide range of highly dexterous tasks in real-time. bidex-teleop.github.io

Can robots perform complex, multi-stage tasks like making breakfast or organizing a cluttered shelf based solely on text instructions? In our latest work ManipGen, we introduce local policy, a novel policy class that focuses on local interaction regions to enhance generalization…

Murtaza Dalal @mihdalal

10 months ago

Can robots perform complex, multi-stage tasks like making breakfast or organizing a cluttered shelf based solely on text instructions? In our latest work ManipGen, we introduce local policy, a novel policy class that focuses on local interaction regions to enhance generalization…

10 118 622 145K 359

Download Video

Human videos have long helped pre-train visual representations for robotics, but now with large-scale real-robot datasets, people can use in-domain data directly. Check out this work to see how it’s done!

Guangqi Jiang @LuccaChiang

10 months ago

Human videos have long helped pre-train visual representations for robotics, but now with large-scale real-robot datasets, people can use in-domain data directly. Check out this work to see how it’s done!

2 24 107 19K 63

Download Video

Can a single neural network policy generalize over poses, objects, obstacles, backgrounds, scene arrangements, in-hand objects, and start/goal states? Introducing Neural MP: A generalist policy for solving motion planning tasks in the real world 🤖 1/N