Deepak Pathak's banner

Deepak Pathak

@deepakpathak • 31,251 subscribers

Co-Founder & CEO @SkildAI, Faculty @CarnegieMellon. PhD @UCBerkeley; BTech @IITKanpur I study topics in AI (robotics, machine learning & computer vision).

Shorts

Incredible news. Neural MP has won the Best Student Paper award at IROS 2025!! Congratulations to Murtaza Dalal & Jiahui(Jim) Yang for leading the project along with Russell Mendonca, youssef, Russ Salakhutdinov. Neural MP is a major step in making motion planning end-to-end, fast & reactive.

Incredible news. Neural MP has won the Best Student Paper award at IROS 2025!! Congratulations to Murtaza Dalal & Jiahui(Jim) Yang for leading the project along with Russell Mendonca, youssef, Russ Salakhutdinov. Neural MP is a major step in making motion planning end-to-end, fast & reactive.

40,828 Aufrufe

Videos

Anya Rossi

sweetdream.ai

SweetDream.ai•Sponsored•Livecam

Watch Anya Live

Anya is streaming live right now! Join her private show and enjoy exclusive content.

Exclusive private shows

1.2k viewers online

Private Show

Join now for exclusive access

Free preview available • Premium content

"Does S1 exhibit physical prompt steerability: different prompts induce distinct behavior in the same environment?" Yes! Watch S1 follow 4 different video prompt recipes in the same kitchen. The first one is something far out of distribution -- "putting a plate in a toaster".

"Does S1 exhibit physical prompt steerability: different prompts induce distinct behavior in the same environment?" Yes! Watch S1 follow 4 different video prompt recipes in the same kitchen. The first one is something far out of distribution -- "putting a plate in a toaster".

282,316 Aufrufe • vor 28 Tagen

If S1 makes a mistake, it tries again. For example, the input video prompt here attached the wheel just once. During deployment the wheel didn't align properly, so S1 re-aligned it. By watching the prompt, S1 understands the goal and improvises to get there:

If S1 makes a mistake, it tries again. For example, the input video prompt here attached the wheel just once. During deployment the wheel didn't align properly, so S1 re-aligned it. By watching the prompt, S1 understands the goal and improvises to get there:

240,102 Aufrufe • vor 27 Tagen

Humanoid football had a big year. For comparison, this is what the world finals of RoboCup 2026 looked like just last month in South Korea. Should we compete next year? ;)

Humanoid football had a big year. For comparison, this is what the world finals of RoboCup 2026 looked like just last month in South Korea. Should we compete next year? ;)

72,816 Aufrufe • vor 13 Tagen

We hosted Prof. Alyosha Efros (UC Berkeley) at Skild AI! He didn't believe that robots could actually cook eggs reliably. :) Tested back-to-back 5times without fail! One batch of scrambled eggs every ~2.5mins nonstop. The same model assembles a GPU on a server rack too.

We hosted Prof. Alyosha Efros (UC Berkeley) at Skild AI! He didn't believe that robots could actually cook eggs reliably. :) Tested back-to-back 5times without fail! One batch of scrambled eggs every ~2.5mins nonstop. The same model assembles a GPU on a server rack too.

182,940 Aufrufe • vor 6 Monaten

At Skild AI, we’ve raised $1.4B, bringing our valuation to over $14B. We’re on a generational mission, and I’m grateful to be working alongside an exceptional team. Thanks to our investors for the long-term conviction towards omni-bodied intelligence 🚀

At Skild AI, we’ve raised $1.4B, bringing our valuation to over $14B. We’re on a generational mission, and I’m grateful to be working alongside an exceptional team. Thanks to our investors for the long-term conviction towards omni-bodied intelligence 🚀

160,454 Aufrufe • vor 8 Monaten

Even after 4yrs of locomotion research, we keep getting surprised by how far we can push the limits of legged robots! We report a major update 🚀🤖 Extreme Parkour: extremely long & high jumps, ramp, handstand, etc. all with a single neural net! 🧵(1/n)

Even after 4yrs of locomotion research, we keep getting surprised by how far we can push the limits of legged robots! We report a major update 🚀🤖 Extreme Parkour: extremely long & high jumps, ramp, handstand, etc. all with a single neural net! 🧵(1/n)

260,585 Aufrufe • vor 3 Jahren

Force is arguably the most overlooked ingredient in modern robot learning. Introducing FACTR 2: it turns *any* commodity robot into a force-aware system with no force sensors required. Train a tiny force network in <1min with <10mins of data and drop it into any existing teleop pipelines: ✅ Free force sensing for both the robot and the operator arm ✅ Makes demos higher-quality → fewer of them needed. ✅ A new force-aware learning algorithm (FIRST) uses those recovered forces to figure out which parts of a demo actually matter, making learning data-efficient. ✅ Strong performance on complex tasks with fewer demos and even no pretraining! More details below.

Force is arguably the most overlooked ingredient in modern robot learning. Introducing FACTR 2: it turns any commodity robot into a force-aware system with no force sensors required. Train a tiny force network in <1min with <10mins of data and drop it into any existing teleop pipelines: ✅ Free force sensing for both the robot and the operator arm ✅ Makes demos higher-quality → fewer of them needed. ✅ A new force-aware learning algorithm (FIRST) uses those recovered forces to figure out which parts of a demo actually matter, making learning data-efficient. ✅ Strong performance on complex tasks with fewer demos and even no pretraining! More details below.

41,055 Aufrufe • vor 3 Monaten

Robots assembling robot brain -- imagine this kind of robustness on every precision manufacturing line! Live demo of GPU rack assembly at #NVIDIAGTC: - end-to-end neural network (Skild Brain) finetuned with little data - memory to perform long horizon task (placing jigs, 16 screwes, removing jigs) - robust to disturbances and fast to set up - no fancy sensors, just off-the-shelf arms and cameras

Robots assembling robot brain -- imagine this kind of robustness on every precision manufacturing line! Live demo of GPU rack assembly at #NVIDIAGTC: - end-to-end neural network (Skild Brain) finetuned with little data - memory to perform long horizon task (placing jigs, 16 screwes, removing jigs) - robust to disturbances and fast to set up - no fancy sensors, just off-the-shelf arms and cameras

46,392 Aufrufe • vor 6 Monaten

Introducing ManipGen: a sim2real agent for zero-shot manipulation. ManipGen handles complex tasks in the real world like organizing shelves, tidying cluttered tables, and more – all from text input and with on human demonstrations!

Introducing ManipGen: a sim2real agent for zero-shot manipulation. ManipGen handles complex tasks in the real world like organizing shelves, tidying cluttered tables, and more – all from text input and with on human demonstrations!

58,160 Aufrufe • vor 1 Jahr

🤖 Robotics often faces a chicken and egg problem: no web-scale robot data for training (unlike CV or NLP) b/c robots aren't deployed yet & vice-versa. Introducing VRB: Use large-scale human videos to train a *general-purpose* affordance model to jumpstart any robotics paradigm!

🤖 Robotics often faces a chicken and egg problem: no web-scale robot data for training (unlike CV or NLP) b/c robots aren't deployed yet & vice-versa. Introducing VRB: Use large-scale human videos to train a general-purpose affordance model to jumpstart any robotics paradigm!

70,743 Aufrufe • vor 3 Jahren

Keine weiteren Inhalte verfügbar