Lucy Cai's banner
Lucy Cai's profile picture

Lucy Cai

@lucyjcai • 1,971 subscribers

building @tryinstance (YC S26) I like robots! @MIT_LISLab MIT

Shorts

1/ we trained a behavior cloning policy on 40 demos of picking and placing a plush toy. the result is so smooth that you would think it was teleoperated. it's also badly overfit. to explore how well it generalizes, we probed it along four axes to find where it suffers most.

1/ we trained a behavior cloning policy on 40 demos of picking and placing a plush toy. the result is so smooth that you would think it was teleoperated. it's also badly overfit. to explore how well it generalizes, we probed it along four axes to find where it suffers most.

24,593 views

pov behind the scenes of collecting training data for a mini pick and place policy

pov behind the scenes of collecting training data for a mini pick and place policy

20,438 views

Videos

LIVE
1.2k

Anya Rossi

sweetdream.ai

SweetDream.aiSponsored•Livecam
Streaming Now

Watch Anya Live

Anya is streaming live right now! Join her private show and enjoy exclusive content.

HD live stream
Exclusive private shows
1.2k viewers online

Free preview available • Premium content

We’re excited to launch Instance – a success detector for robot rollouts! Describe your task and drop in a dataset, and it judges success/fail on each rollout with detailed subtask captions– it’s more accurate than Claude Opus 4.8, at a fraction of the latency. This is just the first step toward our larger goal: an autonomous evaluation rig for robot learning. Today, evaluating a robot policy means humans have to watch the robot roll out, mark success, and reset the scene– we're automating that, starting with the success judge.
1:12
lucyjcai's profile picture

We’re excited to launch Instance – a success detector for robot rollouts! Describe your task and drop in a dataset, and it judges success/fail on each rollout with detailed subtask captions– it’s more accurate than Claude Opus 4.8, at a fraction of the latency. This is just the first step toward our larger goal: an autonomous evaluation rig for robot learning. Today, evaluating a robot policy means humans have to watch the robot roll out, mark success, and reset the scene– we're automating that, starting with the success judge.

Lucy Cai

73,929 views • 1 month ago

first time teleoperating our new Anvil Robotics arm!
0:54
lucyjcai's profile picture

first time teleoperating our new Anvil Robotics arm!

Lucy Cai

28,036 views • 1 month ago

today we're launching automatic captioning at Instance. send us your robot data, and every episode is segmented into captioned subtasks, graded success/fail plus 1-5 on quality and speed. if this could be useful for you, reach out and we'll caption an episode for free!
0:54
lucyjcai's profile picture

today we're launching automatic captioning at Instance. send us your robot data, and every episode is segmented into captioned subtasks, graded success/fail plus 1-5 on quality and speed. if this could be useful for you, reach out and we'll caption an episode for free!

Lucy Cai

14,365 views • 1 month ago

Imagine you told a robot to "find your car keys" in your apartment and it looked around, opened a drawer, and retrieved them for you. As a step towards that, I adapted TiPToP to run on the RBY1 humanoid in our lab! Here's an example instruction it follows: "Put the green block on the blue plate and the yellow block on the magazine." TiPToP helps plan over the right arm + single torso joint, but it's easy to unlock more joints -- even the base wheels -- for more expressive, real-world tasks. Humans find objects without thinking twice. One day, robots will too! 🤖
0:39
lucyjcai's profile picture

Imagine you told a robot to "find your car keys" in your apartment and it looked around, opened a drawer, and retrieved them for you. As a step towards that, I adapted TiPToP to run on the RBY1 humanoid in our lab! Here's an example instruction it follows: "Put the green block on the blue plate and the yellow block on the magazine." TiPToP helps plan over the right arm + single torso joint, but it's easy to unlock more joints -- even the base wheels -- for more expressive, real-world tasks. Humans find objects without thinking twice. One day, robots will too! 🤖

Lucy Cai

14,711 views • 5 months ago

No more content to load