Search Results for author:
Found 14 papers, 0 papers with code
A Survey Analyzing Generalization in Deep Reinforcement Learning
Reinforcement learning research obtained significant success and attention with the utilization of deep neural networks to solve problems in high dimensional state or action spaces.
Detecting Adversarial Directions in Deep Reinforcement Learning to Make Robust Decisions
Learning in MDPs with highly complex state representations is currently possible due to multiple advancements in reinforcement learning algorithm design.
Adversarial Robust Deep Reinforcement Learning Requires Redefining Robustness
Learning from raw high dimensional data via interaction with a given environment has been effectively achieved through the utilization of deep neural networks.
Deep Reinforcement Learning Policies Learn Shared Adversarial Features 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.
Assessing Deep Reinforcement Learning Policies via Natural Corruptions at the Edge of Imperceptibility
We demonstrate that the perceptual similarity distance of the minimal natural perturbations is orders of magnitude smaller than the perceptual similarity distance of the adversarial perturbations to the unperturbed observations (i. e. minimal natural perturbations are perceptually more similar to the unperturbed states than the adversarial perturbations), while causing larger degradation in the policy performance.
Training with Worst-Case Distributional Shift causes Overestimation and Inaccuracies in State-Action Value Functions
The utilization of deep neural networks as function approximators for the state-action value function created a new research area for self learning systems, and made it possible to learn optimal policies from high dimensional state representations.
Detecting Worst-case Corruptions via Loss Landscape Curvature in Deep Reinforcement Learning
The non-robustness of neural network policies to adversarial examples poses a challenge for deep reinforcement learning.
Adversarial Training Blocks Generalization in Neural Policies
Deep neural networks have made it possible for reinforcement learning algorithms to learn from raw high dimensional inputs.
Investigating Vulnerabilities of Deep Neural Policies
For the second approach, we propose a novel method to measure the feature sensitivities of deep neural policies and we compare these feature sensitivity differences in state-of-the-art adversarially trained deep neural policies and vanilla trained deep neural policies.
Non-Robust Feature Mapping in Deep Reinforcement Learning
We conduct several experiments in the Arcade Learning Environment (ALE), and with our proposed feature mapping algorithms we show that while the state-of-the-art adversarial training method eliminates a certain set of non-robust features, a new set of non-robust features more intrinsic to the adversarial training are created.
Adversarially Trained Neural Policies in the Fourier Domain
Reinforcement learning policies based on deep neural networks are vulnerable to imperceptible adversarial perturbations to their inputs, in much the same way as neural network image classifiers.
Bounded Myopic Adversaries for Deep Reinforcement Learning Agents
Adversarial attacks against deep neural networks have been widely studied.
Exploring Transferability of Perturbations in Deep Reinforcement Learning
In this paper we propose a more realistic threat model in which the adversary computes the perturbation only once based on a single state.
Daylight: Assessing Generalization Skills of Deep Reinforcement Learning Agents
Deep reinforcement learning algorithms have recently achieved significant success in learning high-performing policies from purely visual observations.
