Yevgen Chebotar Profile
Yevgen Chebotar

@YevgenChebotar

Followers
2K
Following
101
Media
12
Statuses
42

Robotic foundation models @NVIDIA | VLAs, Reinforcement Learning & more 🤖 Previously @GoogleDeepMind and @Figure_robot

Joined March 2017
Don't wanna be here? Send us removal request.
@YevgenChebotar
Yevgen Chebotar
11 days
Excited to join the NVIDIA GEAR team to help build the next generation of open robotic foundation models!
13
2
151
@YevgenChebotar
Yevgen Chebotar
10 months
We've made great progress on Vision-Language-Action Models for humanoids in our new Helix model! Check out the technical report for more details: https://t.co/muMeO2Log7
Tweet card summary image
figure.ai
Figure was founded with the ambition to change the world.
@Figure_robot
Figure
10 months
Meet Helix, our in-house AI that reasons like a human Robotics won't get to the home without a step change in capabilities Our robots can now handle virtually any household item:
6
4
93
@YevgenChebotar
Yevgen Chebotar
2 years
Congrats everyone, 170+ authors and contributors, great to see the robotic field coming together!
@KarlPertsch
Karl Pertsch
2 years
Our OpenX paper won best paper at ICRA! Congrats to all my co-authors! 🎉🎉 This is an ongoing effort, we recently added new datasets from the community that double the size of the OpenX dataset -- keep 'em coming! :) Check datasets & how to contribute: https://t.co/gX5Ir30s5B
0
1
22
@YevgenChebotar
Yevgen Chebotar
2 years
Some personal updates! Excited to join the team @Figure_robot to help building AI for the robot age! 🤖
25
2
164
@YevgenChebotar
Yevgen Chebotar
2 years
Turns out classification loss works surprisingly well for value-based RL, also some nice gains when used with Q-Transfomer!
@aviral_kumar2
Aviral Kumar
2 years
Super simple code change to get value-based deep RL scale *much* better w/ big models across the board on Atari games, robotic manipulation w/ transformers, LLM + text games, & even Chess! Just use classification loss (i.e., cross entropy), not MSE!! https://t.co/0IHSgN4pBj🧵⬇️
0
3
15
@YevgenChebotar
Yevgen Chebotar
2 years
RT-H learns a hierarchy all the way from high-level tasks through low-level “language motions” to robot actions! ✅ Improved performance and generalization through better data sharing ✅ Automated grounded “bottom-up” labeling ✅ Ability to intervene and correct with language
@suneel_belkhale
Suneel Belkhale
2 years
Is language capable of representing low-level *motions* of a robot? RT-Hierarchy learns an action hierarchy using motions described in language, like “move arm forward” or “close gripper” to improve policy learning. 📜: https://t.co/61KZfQNpkY 🏠: https://t.co/voIvzq5Mek (1/10)
1
3
36
@xiao_ted
Ted Xiao
2 years
Had a great time today with @YevgenChebotar and @QuanVng visiting @USCViterbi to give a talk on “Robot Learning in the Era of Foundation Models”. Slides out soon, packed with works from *just the past 5 months* 🤯 Thanks to @daniel_t_seita for hosting!
1
3
57
@YevgenChebotar
Yevgen Chebotar
2 years
Presenting RT-2 poster at CoRL! https://t.co/fA39xewX5O
Tweet card summary image
robotics-transformer2.github.io
Project page for RT-2
@QuanVng
Quan Vuong
2 years
Pictures taken at RT-2 poster at @DannyDriess requests ; ) @YevgenChebotar We miss you @TianheYu CC @hausman_k
0
2
24
@CSatUSC
USC Thomas Lord Department of Computer Science
2 years
How many @CSatUSC grads does it take to create a breakthrough robot? Three, apparently! Congrats @YevgenChebotar @hausman_k and @ryancjulian who worked on Google DeepMind's revolutionary RT-2 AI model. Find out more ⤵️ https://t.co/D4HqA7d7zH @gauravsukhatme @USCViterbi
Tweet card summary image
viterbischool.usc.edu
USC computer science alumni worked on Google's "first-of-its-kind" robot AI model, the Robotic Transformer 2 (RT-2)
0
3
20
@YevgenChebotar
Yevgen Chebotar
2 years
Exciting times for Robot Learning! 60 datasets from 22 different robots and 21 institutions combined in a single Open-X Embodiment data repository, resulting in over 1 million episodes and improved RT-X models! Amazing and a very important collaboration across the world! 🤖🌐
@QuanVng
Quan Vuong
2 years
RT-X: generalist AI models lead to 50% improvement over RT-1 and 3x improvement over RT-2, our previous best models. 🔥🥳🧵 Project website: https://t.co/GAlvFdqwx5
0
1
11
@YevgenChebotar
Yevgen Chebotar
2 years
Our real robot policies significantly improve upon RT-1 and other baselines when trained on limited amount of human demonstrations by leveraging autonomously collected negatives and dynamic programming properties of Q-learning.
1
0
15
@YevgenChebotar
Yevgen Chebotar
2 years
By using autoregressive Bellman updates, conservative regularization, Monte Carlo and n-step returns, we are able to combine human demonstrations and autonomously collected data to learn multi-task language-conditioned policies from both, successful and failed examples.
1
0
17
@YevgenChebotar
Yevgen Chebotar
2 years
Offline RL strikes back! In our new Q-Transformer paper, we introduce a scalable framework for offline reinforcement learning using Transformers and autoregressive Q-Learning to learn from mixed-quality datasets! Website and paper: https://t.co/SntNYC9Pk3 🧵
8
107
524
@YevgenChebotar
Yevgen Chebotar
2 years
Excited to present RT-2, a large unified Vision-Language-Action model! By converting robot actions to strings, we can directly train large visual-language models to output actions while retaining their web-scale knowledge and generalization capabilities! https://t.co/27sBueV42q
@GoogleDeepMind
Google DeepMind
2 years
Today, we announced 𝗥𝗧-𝟮: a first of its kind vision-language-action model to control robots. 🤖 It learns from both web and robotics data and translates this knowledge into generalised instructions. Find out more: https://t.co/UWAzrhTOJG
0
14
76
@hausman_k
Karol Hausman
3 years
Very exited to announce our largest deep RL deployment to date: robots sorting trash end-to-end in real offices! https://t.co/tTASU1Fmgs (aka RLS) This project took a long time (started before SayCan/RT-1/other newer works) but the learnings from it have been really valuable.🧵
16
161
822
@DannyDriess
Danny Driess
3 years
What happens when we train the largest vision-language model and add in robot experiences? The result is PaLM-E 🌴🤖, a 562-billion parameter, general-purpose, embodied visual-language generalist - across robotics, vision, and language. Website: https://t.co/ouMkeQiGr5
27
510
2K
@hausman_k
Karol Hausman
4 years
Super excited to introduce SayCan ( https://t.co/NWyvPubhmE): 1st publication of a large effort we've been working on for 1+ years Robots ground large language models in reality by acting as their eyes and hands while LLMs help robots execute long, abstract language instructions
18
273
1K
@chelseabfinn
Chelsea Finn
4 years
In offline RL, adding data from other tasks can boost generalization but can surprisingly *hurt* performance We analyze this & develop a conservative data sharing approach to help fix it: https://t.co/kAK7UKPLQu w/ @TianheYu @aviral_kumar2 @YevgenChebotar @hausman_k @svlevine
5
23
162