//
sign in
Profile
by @danabra.mov
Profile
by @dansshadow.bsky.social
Profile
by @jimpick.com
AviHandle
by @danabra.mov
AviHandle
by @dansshadow.bsky.social
AviHandle
by @katherine.computer
EventsList
by @katherine.computer
ProfileHeader
by @dansshadow.bsky.social
ProfileHeader
by @danabra.mov
ProfileMedia
by @danabra.mov
ProfilePlays
by @danabra.mov
ProfilePosts
by @danabra.mov
ProfilePosts
by @dansshadow.bsky.social
ProfileReplies
by @danabra.mov
Record
by @atsui.org
Skircle
by @danabra.mov
StreamPlacePlaylist
by @katherine.computer
+ new component
Profile
Loading...
Loading...
Learning to pick from a moving conveyor belt with a compositional world model! As in, the robot learns purely in its own imagination.
7h
How social media feels nowadays smh my head
also IMO the only way we ever get a Chinese (or any open source) mythos
How is everyone getting such high refusal rates with fable 5
Post training deepseek with a 1k gpu Huawei cluster -- no nvidia -- feels like a serious milestone to me "Intelligence too cheap to meter" cant arrive if there's an nvidia/tsmc bottleneck, even if its better for shareholders
fable tokens spend so fast. i get it now
Yeah basically
8h
1d
8h
3h
1d
2h
The best way to get robust, high-quality robot skill is through reinforcement learning; but RL in the real world or simulation has limitations. Instead, Jiazhi Yang et al. do RL in a learned compositional world model. Learn more in ep 86 of RoboPapers: robopapers.substack.com/p/ep86-rise-...
Chris Paxton
7h
Chris Paxton
Chris Paxton
Chris Paxton
“the robot learns purely in its own imagination”
Chris Paxton
Chris Paxton
cafkafk
4h
Chris Paxton
Grace
Video
The best way to get robust, high-quality robot skill is through reinforcement learning; but RL in the real world or simulation has limitations. Instead, Jiazhi Yang et al. do RL in a learned compositional world model. Learn more in ep 86 of RoboPapers: robopapers.substack.com/p/ep86-rise-...
7h
Chris Paxton