Learning through Model Differentiation

Differentiable Model-Based RL

Approach of reinforcement learning where gradients are backpropagated through a differentiable environment model to directly optimize policies.

← Indietro