Making AI Control Policies More Robust to Perturbations

Original: Robust Policy Optimization via Adversarial Importance Sampling

Source: arxiv.org ↗