Towards Exploiting Geometry and Time for Fast Off-Distribution Adaptation in Multi-Task Robot Learning

We train policies for a base set of pre-training tasks, then experiment with adapting to new off-distribution tasks, using simple architectural approaches for re-using these policies as black-box priors. We find that combining low-complexity target policy classes, base policies as black-box priors, and simple optimization algorithms allows us to acquire new tasks outside the base task distribution, using small amounts of offline training data. [ArXiv URL]