Ananye Agarwal (@anag004) / X

Ananye Agarwal

338 posts

Ananye Agarwal

@anag004

building robot brains @SkildAI | MLD PhD at CMU | Prev CS @ IITD.

Joined March 2012

Pinned
Ananye Agarwal
@anag004
Sep 24, 2025
The models we build rarely take us by surprise. Watching these new capabilities emerge at scale was a rare, unforgettable exception.
Skild AI
@SkildAI
Sep 24, 2025
We built a robot brain that nothing can stop. Shattered limbs? Jammed motors? If the bot can move, the Brain will move it— even if it’s an entirely new robot body. Meet the omni-bodied Skild Brain:
00:00
35K
Ananye Agarwal
@anag004
Aug 6, 2024
Want to scale RL with your shiny new GPU? 🚀 In our ICML24 Oral we find that RL algorithms hit a barrier when data is scaled up. Our new algorithm, SAPG, proposes a simple fix. It scales to 25k envs and solves hard tasks where PPO makes no progress. sapg-rl.github.io 1/n
00:00
64K
Ananye Agarwal
@anag004
Jul 9, 2024
As a founding researcher, I have seen @SkildAI grow exponentially. We changed 3 offices, grew 10x in human (and robot) numbers, and become a unicorn in less than a year. If you want to scale up robotics and work with a cracked team of engineers and scientists, come to @SkildAI.
Deepak Pathak
@pathak2206
Jul 9, 2024
Thrilled to announce @SkildAI! Over the past year, @gupta_abhinav_ and I have been working with our top-tier team to build an AI foundation model grounded in the physical world. Today, we’re taking Skild AI out of stealth with $300M in Series A funding: forbes.com/sites/rashishr…
40K
Ananye Agarwal
@anag004
Dec 21, 2023
Robotic intelligence requires dexterous tool use, but generalizing across tools is hard. Our CoRL23 paper combines semantics (affordances) with low-level control (sim2real) to show functional grasping that generalizes to hammers, drills and more! dexfunc.github.io 1/n
00:00
28K
Ananye Agarwal
@anag004
Oct 1, 2024
As roboticists, we often wish robots were like GPUs - leave them running overnight and get a trained model in the morning. @mendonca_rl did just that, see his project for insights into reset-free RL!
Russell Mendonca
@mendonca_rl
Oct 1, 2024
Our CoRL 2024 paper shows Reinforcement Learning can allow robots to learn skills via real-world practice, without any demonstrations or simulation engineering. Rewards are provided using language/vision models, and mobility of robots enables autonomous exploration. 1/N
00:00
11K
Ananye Agarwal
@anag004
Dec 18, 2022
Thrilled to receive the Best Systems Paper award for our vision-based locomotion work!!
Deepak Pathak
@pathak2206
Dec 18, 2022
Wonderful week at @corl_conf. We carried 3 robots across the world from CMU to New Zealand for live demos of both papers. Congratulations to @anag004 & @ashishkr9311 for winning the Best Systems Paper Award and to @zipengfu & @xuxin_cheng for being one of the 4 finalists!! 😇
00:00
16K
Ananye Agarwal
@anag004
Jul 29, 2025
To me, the most striking thing about this video is the sheer diversity of robots and tasks. This showcases our thesis: AGI that truly understands the physical world will be omni-bodied -- a single brain to control any robot for any task. We will show more soon!
Skild AI
@SkildAI
Jul 29, 2025
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.
00:00
9.5K
Ananye Agarwal
@anag004
Jul 28, 2025
tomorrow 👀
Skild AI
@SkildAI
Jul 28, 2025
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.
00:00
4.2K
Ananye Agarwal
@anag004
Nov 15, 2022
Excited to share our #CoRL2022 oral paper! We train our small, inexpensive, robot dog to traverse stepping stones, climb stairs and rocks. All of this is done without explicit state estimation and with no hardcoded leg motions, optimal gaits emerge via reinforcement learning.
Deepak Pathak
@pathak2206
Nov 15, 2022
After 3yrs of locomotion research, we report a major update in our #CoRL2022 (Oral) paper: vision-based locomotion. Our small, safe, low-cost robot can walk almost any terrain: high stairs, stepping stones, gaps, rocks. Stair for this robot is like climbing walls for humans.
00:00
Ananye Agarwal
@anag004
Sep 26, 2023
End-to-end learning is surprisingly powerful! In our paper (w/ @xuxin_cheng @teenhaci @pathak2206) we do extreme parkour with legged robots where even a single error can be fatal! We were blown away to see a small, low-cost robot jump twice its height, do handstands and more.
Deepak Pathak
@pathak2206
Sep 26, 2023
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! extreme-parkour.github.io 🧵(1/n)
00:00
6K
Ananye Agarwal
@anag004
Aug 6, 2025
It's truly incredible to see a humanoid gracefully climb long staircases like a human! This is all a single neural network, direct from pixels to actions. No mapping. No planning. No manual "stair mode".
Skild AI
@SkildAI
Aug 6, 2025
We’ve all seen humanoid robots doing backflips and dance routines for years. But if you ask them to climb a few stairs in the real world, they stumble! We took our robot on a walk around town to environments that it hadn’t seen before. Here’s how it works🧵⬇️
00:00
2.6K
Ananye Agarwal
@anag004
Feb 14, 2023
Curious robots in the real world! This robot chooses actions which lead to a change in the environment to rapidly discover useful behaviors.
Deepak Pathak
@pathak2206
Feb 14, 2023
After working on curiosity-driven learning for many years, we finally scale it to real robot control. In our #ICRA2023 paper, we present ALAN which explores w/o any rewards to collect its own real-world data & then repurposes its experience to achieve goals at deployment.
00:00
4.6K
Ananye Agarwal
@anag004
May 29, 2024
In our new #CVPR24 paper, we show whole body manipulation in cluttered real-world settings, all from a single end-to-end neural network mapping pixels to torques!
Shagunn
@shagunuppls
May 29, 2024
Want a robot to navigate a cluttered room and fetch you something? Presenting SPIN: seamlessly moves past obstacles using active vision & whole-body coordination. SPIN learns an end-to-end policy in simulation, jointly optimizing perception & action spin-robot.github.io 1/n
00:00
4.5K
Ananye Agarwal
@anag004
Oct 4, 2023
Great to run into Sam and others from Robotis at IROS! LEAP hand parts can be bought with a single click from their website (robotis.us/leap-hand-bund…). None of this would be possible without @kenny__shaw who couldn't attend but was there with us in spirit :)
ROBOTIS AMERICA
@robotisamerica
Oct 3, 2023
2023 IROS Day was an extremely busy day, but we met a lot of awesome people, including Deepak Pathak (@pathak2206) and Ananye Agarwal (@anag004), two of the creators o the LEAP hand platform! Come see us at 2023 IROS Booth 329 at Huntington Place in Detroit until Wednesday 10/4.
20K