C. Zhang w/o shitpost
@chongzhang.bsky.social
Stupid #robotics guy at ETHz
Twitter: https://x.com/ChongZitaZhang
Research Website: https://zita-ch.github.io/
Twitter: https://x.com/ChongZitaZhang
Research Website: https://zita-ch.github.io/
Jemin is literally the one that solves every big problem for legged robots.
Even wilder that our lab is still using the infra he left 7 years ago to publish SciRo-level papers. His new lab also keeps publishing big works, since the first year he started the lab.
Even wilder that our lab is still using the infra he left 7 years ago to publish SciRo-level papers. His new lab also keeps publishing big works, since the first year he started the lab.
May 30, 2025 at 4:27 AM
Jemin is literally the one that solves every big problem for legged robots.
Even wilder that our lab is still using the infra he left 7 years ago to publish SciRo-level papers. His new lab also keeps publishing big works, since the first year he started the lab.
Even wilder that our lab is still using the infra he left 7 years ago to publish SciRo-level papers. His new lab also keeps publishing big works, since the first year he started the lab.
Saved my experience!
March 29, 2025 at 4:43 PM
Saved my experience!
What should I do for a change?
March 29, 2025 at 1:57 AM
What should I do for a change?
Hacking PPO is all you need?
March 29, 2025 at 1:50 AM
Hacking PPO is all you need?
Yeah that one everyone I know got pissd off
February 14, 2025 at 1:46 AM
Yeah that one everyone I know got pissd off
Ahh everyone saw that
February 14, 2025 at 1:38 AM
Ahh everyone saw that
Reposted by C. Zhang w/o shitpost
The key to achieving horizon generalization is *planning invariance*. A policy is planning invariant if decomposing tasks into simpler subtasks doesn't improve performance. We prove planning invariance can enable horizon generalization. 3/
February 4, 2025 at 8:37 PM
The key to achieving horizon generalization is *planning invariance*. A policy is planning invariant if decomposing tasks into simpler subtasks doesn't improve performance. We prove planning invariance can enable horizon generalization. 3/
I'd say that line of work is really impressive even when they don't need a legged robot. Could be more impressive if they don't opt for the spot.
November 20, 2024 at 2:35 PM
I'd say that line of work is really impressive even when they don't need a legged robot. Could be more impressive if they don't opt for the spot.
Nothing different from being a PhD except I have no salary rn
November 12, 2024 at 1:51 AM
Nothing different from being a PhD except I have no salary rn
These are gifs but won't autoplay. This new software seems to still have some bugs
November 11, 2024 at 6:16 PM
These are gifs but won't autoplay. This new software seems to still have some bugs
Yeeh field roboticist here, all papers learning tho
November 11, 2024 at 9:36 AM
Yeeh field roboticist here, all papers learning tho