Adversarial Reinforcement Learning Papers

This is a collection of adversarial reinforcement learning papers. Each category is a potential start point for you to start your research. Some papers are listed more than once because they belong to multiple categories.

Adversarial reinforcement learning is closely related to robust reinforcement learning and attacks in reinforcement learning. If you are looking for papers in adversarial reinforcement learning, you should also see papers related to robust reinforcement learning and attacks in reinforcement learning.

For MARL resources, please refer to Multi Agent Reinforcement Learning papers, MARL Papers with Code and MARL Resources Collection.

I will continually update this repository and I welcome suggestions. (missing important papers, missing categories, invalid links, etc.) This is only a first draft so far and I'll add more resources in the next few months.

This repository is not for commercial purposes.

My email: [email protected]

Overview

• Single-Agent
• Multi-Agent
• Adversarial Communication
• Benchmark

Single-Agent

Paper	Code	Accepted at	Year
Robust Adversarial Reinforcement Learning	Non-official implements on GitHub	ICML	2017
Robust Deep Reinforcement Learning against Adversarial Perturbations on State Observations	https://github.com/chenhongge/StateAdvDRL	NIPS	2020
Robust Reinforcement Learning as a Stackelberg Game via Adaptively-Regularized Adversarial Training			2022
Risk Averse Robust Adversarial Reinforcement Learning		ICRA	2019
Robust Deep Reinforcement Learning with Adversarial Attacks			2017
Robust Reinforcement Learning on State Observations with Learned Optimal Adversary	https://github.com/huanzhang12/ATLA_robust_RL	ICLR	2021
Exploring the Training Robustness of Distributional Reinforcement Learning against Noisy State Observations			2021
RoMFAC: A Robust Mean-Field Actor-Critic Reinforcement Learning against Adversarial Perturbations on States			2022
Adversary Agnostic Robust Deep Reinforcement Learning		TNNLS	2021
Learning to Cope with Adversarial Attacks			2019
Adversarial Attack on Graph Structured Data		ICML	2018
Characterizing Attacks on Deep Reinforcement Learning		AAMAS	2022
Adversarial policies: Attacking deep reinforcement learning	https://github.com/HumanCompatibleAI/adversarial-policies	ICLR	2020
Learning Robust Policy against Disturbance in Transition Dynamics via State-Conservative Policy Optimization		AAAI	2022
On the Robustness of Safe Reinforcement Learning under Observational Perturbations			2022
Robust Reinforcement Learning using Adversarial Populations			2020
Robust Deep Reinforcement Learning through Adversarial Loss	https://github.com/tuomaso/radial_rl_v2	NIPS	2021

Multi-Agent

Paper	Code	Accepted at	Year
Certifiably Robust Policy Learning against Adversarial Communication in Multi-agent Systems			2022
Distributed Multi-Agent Deep Reinforcement Learning for Robust Coordination against Noise			2022
On the Robustness of Cooperative Multi-Agent Reinforcement Learning		IEEE Security and Privacy Workshops	2020
Towards Comprehensive Testing on the Robustness of Cooperative Multi-agent Reinforcement Learning		CVPR workshop	2022
Robust Multi-Agent Reinforcement Learning via Minimax Deep Deterministic Policy Gradient		AAAI	2019
Multi-agent Deep Reinforcement Learning with Extremely Noisy Observations		NIPS Deep Reinforcement Learning Workshop	2018
Policy Regularization via Noisy Advantage Values for Cooperative Multi-agent Actor-Critic methods			2021

Adversarial Communication

Paper	Code	Accepted at	Year
Certifiably Robust Policy Learning against Adversarial Communication in Multi-agent Systems			2022

Benchmark

Paper	Code	Accepted at	Year
Towards Comprehensive Testing on the Robustness of Cooperative Multi-agent Reinforcement Learning		CVPR workshop	2022

Citation

If you find this repository useful, please cite our repo:

@misc{chen2022adversarial,
  author={Chen, Hao},
  title={Adversarial Reinforcement Learning Papers},
  year={2022}
  publisher = {GitHub},
  journal = {GitHub Repository},
  howpublished = {\url{https://github.com/TimeBreaker/Adversarial-Reinforcement-Learning-Papers}}
}