Team Idea
Bimanual VLA manipulation. Thinking of experimenting with smolvla as a baseline and based on compute availability can move upto molmoact 2 & can experiment with S1 gen 1.5 foundational model.
Setting up dinner table requires multi step reasoning. And coordination with both hands. Can fine tune pi0.7 for this specific task.
The biggest problem would be s0101 is janky at times. And making two of them coordinate writing a bimanual policy
Team Members
Unknown Hacker
dev272dv30840
Submission
Team Leader hasn't made a submission yet