Fetching the paper…

RaC: Robot Learning for Long-Horizon Tasks by Scaling Recovery and Correction · Around