Harsh

building agents and poses @nvidia | Robograd Biorobotics lab @CMU_Robotics | Investing in startups | Taking side of folks who build

[Harsh @HSlifelearner]

Check the full suite for full mocap to robotics pretraining . SOMA has anatomically correct joint definitions and has much detailed mesh key points compared to MHR/SMPL. Foundational for all bodypose downstream tasks. More on this soon on its capabilities.

[Umar Iqbal @UmarIqb]

#NVIDIA just released a whole ecosystem for human(oid) motion and robot learning from human data. 🚀🦾

Data, as we all know, is the key to scaling AI models. To accelerate the field of Embodied AI, we have open-sourced a full stack of models and tools to capture, generate, retarget, and simulate human(oid) motion data at scale, along with a massive high-quality dataset and a standard human skeletal representation, SOMA, to make them all seamlessly communicate with each other.

The entire suite is available under the Apache 2.0 license.

1️⃣ SOMA: A universal interface to unify all parametric human body models (SOMA-shape, SMPL, MHR, etc.) into a standard skeletal representation, eliminating the need for custom adapters or model-specific retargeting.
🔗 https://t.co/Xrg672T7Nu

2️⃣ Kimodo: High-fidelity, controllable text-to-motion generation for both humans and humanoid robots.
🔗 https://t.co/2cQKAPfvEU

3️⃣ GEM: A global human pose estimation method from in-the-wild videos, natively compatible with SOMA.
🔗 https://t.co/pV0043jwcO

4️⃣ Bones-SEED: A massive dataset of 150k+ motions in SOMA format, including data already retargeted for the Unitree G1, created with our partners at Bones Studio.
🔗 https://t.co/wxfyZ7S9TJ
🔗 https://t.co/oM5rIMdRi8

5️⃣ SOMA Retargeter: A dedicated tool for seamless motion retargeting from the SOMA skeleton to the Unitree G1.
🔗 https://t.co/jg4DUjWcnw

6️⃣ ProtoMotions: Our high-performance simulation framework for training digital human(oid)s via RL, now with native SOMA support.
🔗 https://t.co/K1zsGEdl5S

This is just the beginning, and we have much more in the pipeline. Excited to see what the community builds next!

#NVIDIA #GTC #GTC2026 #Robotics #EmbodiedAI #PhysicalAI @NVIDIAAI

[Sanjil Jain 🤖/acc @Sanjil]

Simulation is the key to scaling in robotics and CoRL just proved it:

1. MolmoBot
> Simulation only on MuJoCo
> 79.2% zero-shot on real tasks

2.GRAIL
> Humanoid data made from 3D assets and a video model, no teleoperation
> 84% pick-up on a real G1 and about 90% on stairs

3. ZAP
> Open 27B model trained only in simulation
> 96.7% on a real bimanual arm and matched a frontier model at 1/17 of its cost

4. RPG
> 28.6% to 95% in 15 practice rounds

5. DreamZero
> 12 min of human video took unseen tasks from 38% to 54%

6. GR00T 2
> 95% on new tasks from 5 demos
> Most of its training was human video

The best real-world results are now coming from robots that never even touched the real world!

Excited to see what unfolds in the coming days!

[David @DavidSHolz]

@Karmedge no one ever seems to ask "what if we put a billion dollars into simulators instead of data collection"

[Dyna Robotics @DynaRobotics]

Yesterday, we unboxed Taku at @corl_conf put it to work.

Powered by Dyna-2.1, it performed long-horizon, whole-body mobile manipulation zero-shot, with no on-site data collection or post-training.

Come see it live and meet the team at Booth 500 in Griffin Hall!

[Harsh @HSlifelearner]

@chris_j_paxton 👏

[Kevin Wimer @kevinace]

I have stage 4 cancer. I thought I'd fight it alone.

I was wrong.

Since mid October, ~20 strangers on X have stepped up to help me. All for free so far:
- An NYU Langone lab offered to run my tumor data through their system to sharpen my vaccine targets
- A U Michigan lab offered to test drugs on my actual tumor cells
- A biotech ran AI analysis on my tumor RNA
- A vaccine company offered to design a second vaccine to check the first one

Scientists. Engineers. People I've never met.

Cancer has been picking us off one at a time for all of human history. It counts on us fighting it alone, scared and quiet.

Not anymore.

We have AI. We have the internet. We have each other.

Cancer, we're coming for you, bitch. Buckle up!

[NVIDIA Robotics @NVIDIARobotics]

From simulation to the RoboCup semifinals. ⚽

@UTAustin's humanoid soccer team finished fourth among 22 teams in RoboCup 2026’s Large Robot Division.

Before competing, the team developed and tested robot skills in simulation using shared GPUs at the Texas Advanced Computing Center (@TACC) and NVIDIA Isaac.

See how the team took those skills onto the field: https://t.co/68QvKCdxOV

#CoRL2026

[Harsh @HSlifelearner]

@dwarkesh_sp @TheRealAdamG For same team per model , One other eval could be interesting is to have a weaker llm in the team and see how team perform degrades .

[Harsh @HSlifelearner]

@dwarkesh_sp @TheRealAdamG Have you tried mix and match the teams from different models and let a judge evaluate player performance end of the game ?
Same model being on same team might be a bias issue .

[Jan Kautz @jankautz]

Amazing work by the entire NVIDIA robotics team!

[Harsh @HSlifelearner]

😂😂

[Bryan Catanzaro @ctnzr]

So proud that @NVIDIAGeForce DLSS5 has won the 2027 Best of Innovation in Gaming & eSports award at CES, and also is an honoree for Artificial Intelligence as a category.

https://t.co/NEU21agI2J

[Harsh @HSlifelearner]

Groot2 is here !!

[Yuke Zhu @yukez]

Meet GR00T 2, our new generation of world-action models (WAMs): it imagines how the physical world will unfold and generates the actions that get it there. One model runs on parallel grippers, dexterous hands, and humanoids.

https://t.co/32Z1plSGQg

[Yan Chang @Dr_YanChang]

Human → Sim → Real: turning human videos into robot skills.

Join us at CoRL for “From Human Videos to Dexterous Robot Capability: A Hands-on Tutorial”! https://t.co/92R9yMjv6c
📅 Thursday, November 12, 2026 | 2:00–6:00 PM CST
📍 Hilton Austin, Salon D · Austin, Texas

We’ll share a detailed walkthrough of the complete end-to-end video-to-data pipeline that enables “Human → Sim → Real” adaptability for GR00T 2—from human demonstration videos through simulation to training data for real-world robot skills. https://t.co/LwhBs3yfO1

We’ll also feature:
• A panel on the future of robot learning from humans, exploring opportunities and open challenges.
• The announcement of the V2D Challenge winners, followed by presentations from the winning teams on their solutions.
Kudos to the panelists, instructors, and organizers bringing this tutorial to life!

Panelists:
• @Jeannette Bohg — Stanford University @leto__jean
• @Kuan Fang — Cornell University @KuanFang
• Yuting Ye — Meta
• Kaifeng “Alex” Zhang — Sharpa
• @Roei Herzig — UC Berkeley @roeiherzig
• Louis Lian — Lightwheel
• Javier Romero — Meta
• Yan Chang — NVIDIA @Dr_YanChang
Instructors: @zhu_xinghao, @Mrinal_Verghese, @bowenwen_me, Zeo Liu, @shalin_sj, @Naema Bhatti, and @_abraranwar .

Organizers: Xinghao Zhu, Yan Chang, Zixi (Zeo) Liu, Bowen Wen, Mrinal Verghese, Shalin Jain, John Welsh, Daniel Zou, Abrar Anwar, Naema Bhatti, Michael Lin, Kuan Fang, and Joydeep Biswas.

Thank you for sharing your expertise and making this hands-on learning experience possible!
Explore the agenda and join us: CoRL tutorial details.
#CoRL2026 #RobotLearning #PhysicalAI #GR00T #VideoToData #NVIDIA

[Harsh @HSlifelearner]

Nothing to see here

[Claude @claudeai]

Introducing Claude Fable 5.2 and Claude Mythos 5.2, designed to tackle the most daunting technical challenges.

They’re capable collaborators on ambitious projects in many fields, and show improvements in areas like coding, reasoning, and vision.

[Zhengyi “Zen” Luo @zhengyiluo]

@ Austin for #CORL2026! Let's catch up!

We will be unveiling Gr00t 2 tomorrow and presenting GRAIL, CHIP, and CHORD!

[Hynek Kydlíček @HKydlicek]

SOMA-X is actually one of the biggest gift from nvidia to open vision research. One no longer has to fear of license issues with simplx and MANO.

There are also much better

[Jim Fan @DrJimFan]

CoRL Keynote. Hilton Austin. Monday Nov 9, 1:45 pm. 2 days countdown.

Something extraordinary.

[Yuke Zhu @yukez]

Something exciting is coming next Monday 👀

[Chris Paxton @chris_j_paxton]

CoRL has become the place to be if you are interested in robotics. There's no robotics without intelligence -- robots will never scale without scalable learning for problems like contact-rich mobile manipulation -- and so we've seen this academic conference become so essential IMO. Looking forward to meeting people there