Deep Reinforcement Learning Policies Learn Shared Adversarial Features Across Mdps

Abstract

The use of deep neural networks as function approximators has led to striking progress for reinforcement learning algorithms and applications. Yet the knowledge we have on decision boundary geometry and the loss landscape of neural policies is still quite limited. In this paper we propose a framework to investigate the decision boundary and loss landscape similarities across states and across MDPs. We conduct experiments in various games from Arcade Learning Environment, and discover that high sensitivity directions for neural policies are correlated across MDPs. We argue that these high sensitivity directions support the hypothesis that non-robust features are shared across training environments of reinforcement learning agents. We believe our results reveal fundamental properties of the environments used in deep reinforcement learning training, and represent a tangible step towards building robust and reliable deep reinforcement learning agents.

Stats

• citations7
• S2 citations—
• github stars0
• HF likes0
• heat score6.77
• arxiv keykorkmaz2021deep

Related papers

• Detecting Adversarial Directions In Deep Reinforcement Learning To Make Robust Decisions (2023)0.00
• Investigating Vulnerabilities Of Deep Neural Policies (2021)0.00
• Adversarial Policies: Attacking Deep Reinforcement Learning (2019)0.00
• Data Efficient Training For Reinforcement Learning With Adaptive Behavior Policy Sharing (2020)0.00
• Robust Model-based Reinforcement Learning With An Adversarial Auxiliary Model (2024)0.00
• Neutral Agent-based Adversarial Policy Learning Against Deep Reinforcement Learning In Multi-party Open Systems (2025)0.00
• Safe And Robust Experience Sharing For Deterministic Policy Gradient Algorithms (2022)0.00
• Rethinking Adversarial Attacks In Reinforcement Learning From Policy Distribution Perspective (2025)5.84