Robot Learning #02: World Models and Dexterous Manipulation

A builder-focused robotics evening on how data drives robot learning: what models can learn from video and demonstrations, how that learning connects to control, and how progress is evaluated. Topics on the Luma page include how world representations and prediction relate to planning and action, learning manipulation from demonstrations and where transfer to physical tasks falls short, and building training and evaluation datasets and analyzing failure cases. Doors open at 5:00 PM, short demos start at 5:45, a panel with audience Q&A at 6:15, and networking at 7:00, ending about 8:30 PM. Held at Bright Data's Web Data Loft, 625 2nd St. Free, but registration needs host approval.