Johnson Wang
@johnsonwang0810
Followers
319
Following
10
Media
0
Statuses
15
MIT EECS PhD
Massachusetts, USA
Joined April 2015
Introducing GEN-0, our latest 10B+ foundation model for robots โฑ๏ธ built on Harmonic Reasoning, new architecture that can think & act seamlessly ๐ strong scaling laws: more pretraining & model size = better ๐ unprecedented corpus of 270,000+ hrs of dexterous data Read more ๐
48
287
1K
Today, Weโre launching Genesis AI โ a global physical AI lab and full-stack robotics company โ to build generalist robots and unlock unlimited physical labor. Weโre backed by $105M in seed funding from @EclipseVentures, @khoslaventures, @Bpifrance, HSG, and visionaries
22
95
425
Weโre excited to share some updates on Genesis since its release: 1. We made a detailed report on benchmarking Genesis's speed and its comparison with other simulators ( https://t.co/muua5zM71k) 2. Weโve launched a Discord channel and a WeChat group to foster communications
github.com
A Detailed Performance Benchmark Comparison on Genesis vs Isaac Gym & MJX - zhouxian/genesis-speed-benchmark
8
56
323
Soft robot in Genesis!! Hope more ppl can have fun with soft robotics
A cool demo showing soft robot simulation and control in Genesis (yes it's simulated), and then in real world :) made possible by amazing @johnsonwang0810
2
0
11
Welcome to Genesis. Very exciting project to advance robotics and beyond via generative physical simulation.
Everything you love about generative models โ now powered by real physics! Announcing the Genesis project โ after a 24-month large-scale research collaboration involving over 20 research labs โ a generative physics engine able to generate 4D dynamical worlds powered by a physics
0
1
11
[1/4] ๐จ Excited to introduce Embodied Red Teaming (ERT) โ an approach based on vision-language models (VLM) to automatically red team your favorite robotic foundation models!
1
5
12
๐ Introducing NormalFlow: the state-of-the-art tactile tracking algorithm! Just by touch ๐, it precisely tracks objectsโeven with minimal textureโand can even reconstruct objects! ๐ project page: https://t.co/exvt2MFfya paper (RA-L): https://t.co/fIZ3JLZAyN โฌ๏ธ(1/4)
4
20
59
Today we introduce Liquid Foundation Models (LFMs) to the world with the first series of our Language LFMs: A 1B, 3B, and a 40B model. (/n)
86
252
2K
LLM-generated plans in abstract spaces lack physical grounding. Human demonstrations in high-dimensional continuous spaces lack labels for motion constraints. MIT CSAIL-led research grounds language plans in demonstrations as โmanipulation modesโ to build robust policies:
1
23
66
(1/4) ๐ Excited to share our ICLR'24 paper on "Curiosity-driven Red-teaming for Large Language Models"! We bridge curiosity-driven exploration in reinforcement learning (RL) with red-teaming, introducing the Curiosity-driven Red-teaming (CRT) method. #ICLR24 #AI #LLMSecurity
5
7
29
#NeurIPS23 Even wonder when Generative AI be useful for designing real-world robots. Excited to share our oral paper DiffuseBot! Diffusion models and Differentiable Physics are all you need! ๐๐๐๐ Project page: https://t.co/khhhAIpUOK
2
19
108
This is happening tomorrow with an amazing list of speakers and panelists! Join us in Sequoia 1 from 8:30 to 17:30, or via zoom:
cmu.zoom.us
Zoom is the leader in modern enterprise cloud communications.
๐คHow far are we from ๐ ๐๐ง๐๐ซ๐๐ฅ๐ข๐ฌ๐ญ ๐ซ๐จ๐๐จ๐ญ๐ฌ? ๐๐ง๐ง๐จ๐ฎ๐ง๐๐ข๐ง๐ the 1st Workshop on "Towards Generalist Robots" at #CoRL2023! Join us to discuss how to scale up robotic skill learning, with an amazing lineup of speakers! CfP: https://t.co/bPLoeTvwwD Details ๐
3
7
34
Can GPTs generate infinite and diverse data for robotics? Introducing RoboGen, a generative robotic agent that keeps proposing new tasks, creating corresponding environments and acquiring novel skills autonomously! code: https://t.co/PuU2d3WMEs ๐๐งต (better with audio)
11
82
311
๐คHow far are we from ๐ ๐๐ง๐๐ซ๐๐ฅ๐ข๐ฌ๐ญ ๐ซ๐จ๐๐จ๐ญ๐ฌ? ๐๐ง๐ง๐จ๐ฎ๐ง๐๐ข๐ง๐ the 1st Workshop on "Towards Generalist Robots" at #CoRL2023! Join us to discuss how to scale up robotic skill learning, with an amazing lineup of speakers! CfP: https://t.co/bPLoeTvwwD Details ๐
3
36
128
For me, this is the first work featuring event-vision after @davsca1 sparked my interest back in Zurich. The paper has been a cool collaboration with @xanamini, @johnsonwang0810 (co-lead), @WilkoSchwarting, @zhijianliu1996, @SongHan_MIT, @SertacKaraman, and Daniela Rus. 3/n
1
1
4