Post-Hoc Robustness for Model-Based Reinforcement Learning
For practitioners deploying RL agents in safety-critical or adversarial settings, this method offers a way to enhance robustness without retraining, though it is incremental over existing adversarial training approaches.
This work introduces a post-hoc robustification method for model-based RL that improves robustness at inference time without additional neural network training, achieving significant robustness gains in perturbed MuJoCo environments.
To improve the real-world applicability of reinforcement learning (RL), the field of adversarially robust RL studies how to train agents under adversarial environment perturbations. In this setting, a protagonist agent optimizes a policy under environmental perturbations from an adversary, resulting in a zero-sum Markov game. When adversarially robust RL is combined with model-based RL, the adversary can target a learned transition model instead of the training environment. Extending this idea, this work introduces post-hoc robustification of deep RL agents at inference time. By using the learned model in combination with a trained nominal policy, our approach performs a robust policy improvement step. The goal is to improve robustness without any additional training of neural networks. Specifically, we utilize model-predictive control under adversarial rollouts, which are approximated via projected gradient descent within a bounded uncertainty set. Furthermore, these offline rollouts are performed while considering and mitigating out-of-distribution issues. The proposed methodology is validated by demonstrating significant improvements in robustness when the algorithm is evaluated in perturbed Gymnasium MuJoCo environments, while considering the computational limitations of the post-hoc inference setting.