Pinned Tweet
Our new work, STITCH 2.0, can perform consecutive running sutures to close a sample wound with the daVinci robot.
8
15
63
27,233
Kush Hari retweeted
Check out our #CoRL2026 paper COWBOY 🤠!
We learn mobile manipulation from sim-to-real that generalizes to diverse objects and scenes
We learn mobile manipulation from sim-to-real that generalizes to diverse objects and scenes
Context isn’t just what you put in a prompt. It’s also what stands between a robot and its goal. Robot policies in everyday environments must generalize not only across objects, but also across the spatial contexts in which they appear.
Excited to share COWBOY 🤠 (CoRL 2026) — for contextual whole-body manipulation. One generalist grasping policy coordinates base, arm, dexterous hand, and active vision end-to-end, adapting its motion and grasps to object and scene geometry. Trained in simulation, deployed zero-shot in the wild.
w/ @JasonJZLiu @LiuMuxinl
🧵(1/N)
0
6
31
1,160
Kush Hari retweeted
Last year I trained the advantage-conditioned VLA for π*0.6 during my internship @physical_int. Now, I have a much improved recipe.
Introducing ✨Advantage Guidance✨, a performant RL method with stable advantage-conditioned supervised training and contrastive guidance sampling
Introducing ✨Advantage Guidance✨, a performant RL method with stable advantage-conditioned supervised training and contrastive guidance sampling
10
66
588
59,421
Kush Hari retweeted
We gave Claude and GPT a phone video of a robot wiping. In an hour, each built a full-dynamics digital twin.
Introducing RoboSimStudio 🤖: an open and interactive toolkit for Agentic Robotics, enabling Fable, Astra and Argon AI agents to model sponges 🧽, cables 🧵 and cloth 👔.
Introducing RoboSimStudio 🤖: an open and interactive toolkit for Agentic Robotics, enabling Fable, Astra and Argon AI agents to model sponges 🧽, cables 🧵 and cloth 👔.
40
50
439
169,852
Kush Hari retweeted
Want to know how well agents can really control robots directly?
In RoboARCade, we systematically evaluate and diagnose frontier agents controlling robots directly using tens of thousands of dollars’ worth of agent trajectories. RoboARCade contains diverse, precise, and er, code, assets and every trace are out today. roboarcade.org
In RoboARCade, we systematically evaluate and diagnose frontier agents controlling robots directly using tens of thousands of dollars’ worth of agent trajectories. RoboARCade contains diverse, precise, and er, code, assets and every trace are out today. roboarcade.org
10
42
171
16,984
Kush Hari retweeted
(1/N) Who says you can't do contact-rich, precise manipulation with sim-to-real RL?
No demonstrations, no retargeting, no tactile sensors, no crazy reward tuning. Just 1x, reactive policies trained with scaled up RL.
Excited to finally release Success-Guided Sampling (SGS)! g's thread and more thoughts of my own below! 🧵
Website here: sgs-rl.github.io
No demonstrations, no retargeting, no tactile sensors, no crazy reward tuning. Just 1x, reactive policies trained with scaled up RL.
Excited to finally release Success-Guided Sampling (SGS)! g's thread and more thoughts of my own below! 🧵
Website here: sgs-rl.github.io
(1/5) Everyone says dexterous manipulation needs real-world data. These policies were trained with none. Our robots mesh gears, thread nuts, and insert pegs zero-shot from RGB, and the same recipe works for locomotion across terrains much harder than before.
The missing piece rithm, but a small tweak to what the robot practices.
Today we're releasing Success-Guided Sampling (SGS), a simple change to the task sampler that unlocks new levels of dexterity and agility with large-scale, end-to-end sim2real RL.
No demonstrations, no per-task reward tuning. Just PPO at scale.
Check out the policies yourself at sgs-rl.github.io! All videos at 1x speed! 🧵 🔈
13
39
320
34,418
Kush Hari retweeted
Real2sim has made impressive progress with coding agents. But what about sim2real, especially the visual gap? Meet RoboRender: turning simulated trajectories into photorealistic videos for policy training, achieving 71% average success in zero-shot real-world robot deployment.
14
25
281
23,680
Kush Hari retweeted
Very excited to release EyeRobot 2.0! We also built a simulation environment with 6 manipulation tasks and around 1,800 evaluation trials. Demos were collected in VR with the same GELLO-style setup we use on the real robot, and two tasks (pot and tape) mirror our real-world ones. faster and test more ideas at scale, and the same overall trends carried over to the real robot.
Almost all bimanual robots include wrist cameras. They add bulk, cost, and complexity, but are critical for performance. Are they really needed?
Excited to finally share EyeRobot 2.0, where we use active gaze from an ego stereo camera instead 🧵 (with @KushtimusPrime)
3
5
37
1,559
Kush Hari retweeted
Excited to introduce QF3: Fast Flow RL with Filtered Q-Gradients!
Train humanoids from scratch, fine-tune VLAs, even fine-tune image models... all with one simple off-policy update ✨
Train humanoids from scratch, fine-tune VLAs, even fine-tune image models... all with one simple off-policy update ✨
13
59
424
39,315
Excited to release EyeRobot 2.0! We noticed it is better at dealing with distractors. When the gaze policy looks at the right target, the policy can recover
Almost all bimanual robots include wrist cameras. They add bulk, cost, and complexity, but are critical for performance. Are they really needed?
Excited to finally share EyeRobot 2.0, where we use active gaze from an ego stereo camera instead 🧵 (with @KushtimusPrime)
3
10
90
6,924
Kush Hari retweeted
Humans generalize from rich interaction experience. We bring this to humanoid loco-manipulation through scalable real-to-sim-to-real.
Introducing PRISM, we expand 4 real videos into 256 counterfactual variants, recover robot-object trajectories, and train one policy that rse objects, scales, and spatial layouts.
CoRL’26 🚀 Website for details 👉 prism-real2sim2real.github.io
Work done at Amazon FAR w/ @zhenkirito123 @pabbeel @rocky_duan @JitendraMalikCV @carlo_sferrazza @ckarenliu @GuanyaShi @akanazawa
Introducing PRISM, we expand 4 real videos into 256 counterfactual variants, recover robot-object trajectories, and train one policy that rse objects, scales, and spatial layouts.
CoRL’26 🚀 Website for details 👉 prism-real2sim2real.github.io
Work done at Amazon FAR w/ @zhenkirito123 @pabbeel @rocky_duan @JitendraMalikCV @carlo_sferrazza @ckarenliu @GuanyaShi @akanazawa
12
77
372
154,504
Kush Hari retweeted
Can robots help in data centers? It’s tricky when all the cables are the same color.
Introducing TRACE: bi-directional tracing + interactive perception for monochrome cables, @ #IROS2026 from @AUTOLab_Cal @UCBerkeley (1/8)
Introducing TRACE: bi-directional tracing + interactive perception for monochrome cables, @ #IROS2026 from @AUTOLab_Cal @UCBerkeley (1/8)
3
9
28
13,055
Kush Hari retweeted
Happy to announce the full ABC release! We’re also excited that ABC was accepted to CoRL 2026!
Check out our website for code, 400+ hours of sim data on 24 tasks, and 5,850 labeled policy-evaluation episodes.
@arthurallshire @Cinnabar233 @ritvik_singh9 @redstone_hong @davidrmcall
Check out our website for code, 400+ hours of sim data on 24 tasks, and 5,850 labeled policy-evaluation episodes.
@arthurallshire @Cinnabar233 @ritvik_singh9 @redstone_hong @davidrmcall
6
33
259
77,381
Kush Hari retweeted
Grounded API is live.
- SOTA on hand-tracking benchmarks (< 1 cm)
- SOTA on SLAM benchmarks
- In-the-wild ego data -> enriched data in minutes
- Integration with @huggingface @LeRobotHF & @rerundotio
- Built for @BitRobotNetwork RoboCap suite
Technical report & more↓
- SOTA on hand-tracking benchmarks (< 1 cm)
- SOTA on SLAM benchmarks
- In-the-wild ego data -> enriched data in minutes
- Integration with @huggingface @LeRobotHF & @rerundotio
- Built for @BitRobotNetwork RoboCap suite
Technical report & more↓
48
81
459
175,902
Kush Hari retweeted
Our social systems and institutions confuse me all the time. 😕
AI may make them even more complicated. What if it also represents us inside them?
We already delegate representation constantly - your employer negotiates your insurance, your lawyer negotiates your contract, your representatives advocate for your interests.
Wrote up some early thinking on what financial markets can teach us here, what we found in our recent paper Cooperate to Compete (now accepted to EMNLP), and why one of the hardest open problems may be designing the structures agents negotiate inside rather than the agents themselves.
Read my first research blog below!! I would love to hear any thoughts on how folks are thinking about this.
Blog: abbykoneill.github.io/posts/represen…
AI may make them even more complicated. What if it also represents us inside them?
We already delegate representation constantly - your employer negotiates your insurance, your lawyer negotiates your contract, your representatives advocate for your interests.
Wrote up some early thinking on what financial markets can teach us here, what we found in our recent paper Cooperate to Compete (now accepted to EMNLP), and why one of the hardest open problems may be designing the structures agents negotiate inside rather than the agents themselves.
Read my first research blog below!! I would love to hear any thoughts on how folks are thinking about this.
Blog: abbykoneill.github.io/posts/represen…
0
6
22
1,750
Kush Hari retweeted
here’s an uncut, real-time video of our policy stacking and destacking a cup tower
Excited to share SPD: simulation pre-training for dexterity. We pre-trained a policy in simulation and fine-tuned with less than 2 hours of real data (with @sarthakkamat)
3
7
80
9,350
Kush Hari retweeted
Excited to share SPD: simulation pre-training for dexterity. We pre-trained a policy in simulation and fine-tuned with less than 2 hours of real data (with @sarthakkamat)
17
71
539
139,715
Kush Hari retweeted
How can generalist policies adapt to new challenges at deployment using skills they already have?
We optimize VLA *prompt inputs* with reinforcement learning, enabling efficient real-robot adaptation on complex tasks where existing methods struggle. 🧵
semantic-action-rl.github.io
We optimize VLA *prompt inputs* with reinforcement learning, enabling efficient real-robot adaptation on complex tasks where existing methods struggle. 🧵
semantic-action-rl.github.io
3
20
87
20,565
Kush Hari retweeted
Introducing WARP-RM: A Warp-Augmented Relative Progress Reward Model for Data Curation 🧵
We gave our t-shirt folding robot more demonstrations and it got worse. Every extra demo ended in a successfully folded shirt. The data wasn't bad. It was noisy. The policy couldn't tell dead time, and it imitated both equally. So which moments of a demo are actually worth copying?
🌐 Project Website: uynitsuj.github.io/warp-rm
📄 Paper: arxiv.org/abs/2606.28320
💻 Code: github.com/uynitsuj/WARP-…
📨 XDOF blog post: xdof.ai/blog/warp-rm
We gave our t-shirt folding robot more demonstrations and it got worse. Every extra demo ended in a successfully folded shirt. The data wasn't bad. It was noisy. The policy couldn't tell dead time, and it imitated both equally. So which moments of a demo are actually worth copying?
🌐 Project Website: uynitsuj.github.io/warp-rm
📄 Paper: arxiv.org/abs/2606.28320
💻 Code: github.com/uynitsuj/WARP-…
📨 XDOF blog post: xdof.ai/blog/warp-rm
8
25
174
58,775
Kush Hari retweeted
we started a company!!
so, we’re tackling continual learning: what’s the learning algorithm to take arbitrary data — documents, conversations, the models’ own experience — and make better models? how do we scale compute in the same way we’ve already seen with pre-training and inference time, but scaling on the same data we see as humans, day after day with no labels, no rewards?
A lot of the ingredients are out there already (rl, distillation, long-context, sparse / param-efficient architectures, etc.). our team is at the frontier of these topics, and we’re singularly focused on this. we want to understand this problem better than anyone else in the world.
nobody’s solved this problem yet, but even today it’s extremely greenfield opportunity to co-develop research & useful products. in our space, how people interact with the models defines what the data distribution is - and working on this problem end-to-end, from core science to end user, gives us incredible freedom to define the problem and imagine new kinds of experiences.
i expect we’ll use models that continually learn much differently than we’re using them today. it’ll feel different when the models _just know_, and build on our thinking and direction in ways we can’t even imagine. we don’t even know the queries we’re not asking, the things we would do but aren’t able to today.
i’m so excited to share what we’re doing with the world in the coming months!!
and the team is extremely cracked :) tackling this grand challenge and working alongside @jxmnop @EyubogluSabri @dan_biderman @MayeeChen @__howardchen @shizhehe and many others has made every day so fun. come work with us!
so, we’re tackling continual learning: what’s the learning algorithm to take arbitrary data — documents, conversations, the models’ own experience — and make better models? how do we scale compute in the same way we’ve already seen with pre-training and inference time, but scaling on the same data we see as humans, day after day with no labels, no rewards?
A lot of the ingredients are out there already (rl, distillation, long-context, sparse / param-efficient architectures, etc.). our team is at the frontier of these topics, and we’re singularly focused on this. we want to understand this problem better than anyone else in the world.
nobody’s solved this problem yet, but even today it’s extremely greenfield opportunity to co-develop research & useful products. in our space, how people interact with the models defines what the data distribution is - and working on this problem end-to-end, from core science to end user, gives us incredible freedom to define the problem and imagine new kinds of experiences.
i expect we’ll use models that continually learn much differently than we’re using them today. it’ll feel different when the models _just know_, and build on our thinking and direction in ways we can’t even imagine. we don’t even know the queries we’re not asking, the things we would do but aren’t able to today.
i’m so excited to share what we’re doing with the world in the coming months!!
and the team is extremely cracked :) tackling this grand challenge and working alongside @jxmnop @EyubogluSabri @dan_biderman @MayeeChen @__howardchen @shizhehe and many others has made every day so fun. come work with us!
71
53
675
117,830
Kush Hari retweeted
Introducing Do as I Do 👀, a framework to transform everyday human videos into 100s of dexterous robot demos. Co-led with @bhawna_paliwal_ and @HarithejaE, and check out @notmahi's thread!
Here’s a little preview of our dexterous manipulation results. More about how we produce ructions in this mini-thread! 🧵
x.com/notmahi/status…
Here’s a little preview of our dexterous manipulation results. More about how we produce ructions in this mini-thread! 🧵
x.com/notmahi/status…
Robots are the bottleneck in scaling robotics, and learning from human video promises to solve it. But how can chaotic human data ever measure up to sanitized, lab-made teleoperation data?
Introducing Do as I Do: establishing a much needed correspondence between human videos and ome fun insights below: 🧵
8
36
225
28,917