Annotating and processing 15k episodes (5.7+ days, 200+ tasks) of open-source robot eval datasets containing failures For safe deployment, we must cheaply understand failures and unexpected behavior

We will be slowly opensourcing our entire robotics deployment infrastructure on our open source repo we realized that a full-stack platform for robotics is best kept open for teams to integrate and build upon and that doing this is most aligned with our mission of spreading physical AI github.com/interlatent/inter…

first forward deployment at @almond_robotics cloud inference and evals working through interlatent on Axol super powerful robot, first time i felt like a robot could hurt me