Abstract: First-order Policy Gradient (FoPG) algorithms such as Backpropagation through Time and Analytical Policy Gradients leverage local simulation physics to accelerate policy search, ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results