Karl Pertsch

This account is verified. Learn more

@KarlPertsch

Robot Foundation Models @physical_int

kpertsch.github.io Joined July 2015

278 Following 4.8K Followers 459 Posts

Latest posts 5

Pinned post

Karl Pertsch @KarlPertsch Mar 3

This one has been a long time coming: today we’re introducing MEM, an approach for giving VLAs short-term and long-term memory.

Memory is such an obvious capability, but adding it isn’t easy (most VLAs today are memory-less). A short thread on challenges, solutions, and the new capabilities MEM unlocks for us.

2:16

Karl Pertsch @KarlPertsch Jun 15

Over the last few weeks we have seen some evidence of benchmark manipulation on RoboArena, in part flagged by members of the robotics community. We take this very seriously and have taken steps to protect the integrity of the benchmark and ensure fair, unbiased evaluations.
Most importantly, going forward only organizations without an active policy on the leaderboard will be permitted to submit evaluations. We have applied these changes retroactively across the period of suspicious benchmark activity, to make sure all evals on the leaderboard can be trusted.
A big thank you to all the RoboArena evaluators who volunteer their time to provide unbiased robot evaluations, and to the community members who helped audit the evaluation results and flagged potential wrong-doing.

Pranav Atreya @pranav_atreya Jun 13

We observed evidence of benchmark hacking on RoboArena since April. We have taken steps to prevent this in the future, and we have rolled back evals in accordance with these steps to retain the integrity of the benchmark. Read more about our changes here: robo-arena.github.io.
View quoted post on X

Karl Pertsch @KarlPertsch Apr 16

Happy to share some new results! π0.7 comes with memory, and algorithmic advances to pull out more performance and generalization from diverse training data! Check it out!

Physical Intelligence Verified account @physical_int Apr 16

Our newest model, π0.7, has some interesting emergent capabilities: it can control a new robot to fold shirts for which we had no shirt folding data, figure out how to use an appliance with language-based coaching, and perform a wide range of dexterous tasks all in one model!

View quoted post on X

Karl Pertsch @KarlPertsch Feb 26

Use DROID data and PolaRiS sim evals to test your ideas on strong generalist policies!
Congrats to the CoVer-VLA team!

Jacky Kwok Verified account @jackyk02 Feb 24

🧵(6) DROID Eval
CoVer-VLA achieves 14% gains in task progress and 9% in success rate on the challenging red-team PolaRiS benchmark. In the pan cleaning task, π₀.₅ shows incorrect intent, grasping the pan handle. In contrast, CoVer-VLA correctly uses sponge to scrub the pan.

View quoted post on X

Karl Pertsch @KarlPertsch Feb 25

Very exciting to see first steps of our models doing useful things in the world!
Thanks to Ultra and Weave for being great partners in these deployments!

Physical Intelligence Verified account @physical_int Feb 25

General-purpose AI models are behind some of the most exciting applications we now can't live without. We envision that an analogous “physical intelligence layer” built with models like π0.6 will similarly spur a new wave of applications for the physical world.

We’ve recently begun working with a handful of companies that have deployed their robots to do real-world, useful things.

pi.website/blog/partner/?…
View quoted post on X