Adversarial Policy

Adversarial policy research focuses on designing and deploying agents that act as adversaries to test and improve the robustness of other AI systems, particularly in reinforcement learning. Current research emphasizes developing sophisticated adversarial policies using techniques like reinforcement learning (e.g., DDPG, PPO) and generative adversarial networks (GANs), often incorporating elements like uncertainty estimation and human-like risk assessment to create more realistic and effective attacks. This work is crucial for enhancing the safety and reliability of AI systems in safety-critical applications such as autonomous vehicles and human-robot interaction, as well as for understanding and mitigating vulnerabilities in various AI models.

Papers

October 7, 2024

ResTNet: Defense against Adversarial Policies via Transformer in Computer Go
Tai-Lin Wu, Ti-Rong Wu, Chung-Chin Shih, Yan-Ru Ju, I-Chen Wu
Transformer Based Adversarial Policy AlphaZero Based Artificial Intelligence Go Game AlphaZero Algorithm MI Attack

September 19, 2024

VCAT: Vulnerability-aware and Curiosity-driven Adversarial Training for Enhancing Autonomous Vehicle Robustness
Xuan Cai, Zhiyong Cui, Xuesong Bai, Ruimin Ke, Zhenshu Ma, Haiyang Yu, Yilong Ren
Autonomous Vehicle Adversarial Vulnerability Training Time Attack Adversarial Policy Robust Autonomous Driving Curiosity Driven Learning

June 5, 2024

FREA: Feasibility-Guided Generation of Safety-Critical Scenarios with Reasonable Adversariality
Keyu Chen, Yuheng Lei, Hao Cheng, Haoran Wu, Wenchao Sun, Sifa Zheng
Autonomous Vehicle Adversarial Scenario Safety Critical Scenario Adversarial Policy Critical Scenario Generation

March 8, 2024

Overcoming Reward Overoptimization via Adversarial Policy Optimization with Lightweight Uncertainty Estimation
Xiaoying Zhang, Jean-Francois Ton, Wei Shen, Hongning Wang, Yang Liu
Reinforcement Learning Uncertainty Quantification Adversarial Policy Reward Overoptimization

February 6, 2024

SUB-PLAY: Adversarial Policies against Partially Observed Multi-Agent Reinforcement Learning Systems
Oubo Ma, Yuwen Pu, Linkang Du, Yang Dai, Ruo Wang, Xiaolei Liu, Yingcai Wu, Shouling Ji
Multi Agent Reinforcement Learning Multi Agent Target Object Two Player Adversarial Policy

November 17, 2023

From Principle to Practice: Vertical Data Minimization for Machine Learning
Robin Staab, Nikola Jovanović, Mislav Balunović, Martin Vechev
Machine Learning General Principle Empirical Privacy Adversarial Policy Data Minimization

October 30, 2023

Posterior Sampling for Competitive RL: Function Approximation and Partial Observation
Shuang Qiu, Ziyu Dai, Han Zhong, Zhaoran Wang, Zhuoran Yang, Tong Zhang
Posterior Sampling Markov Game Function Approximation Self Play Partial Observation Adversarial Policy

October 16, 2023

Quantifying Assistive Robustness Via the Natural-Adversarial Frontier
Jerry Zhi-Yang He, Zackory Erickson, Daniel S. Brown, Anca D. Dragan
Native Robustness Robust Policy Adversarial Behavior Adversarial Policy Natural Adversarial Collaborative Reinforcement Learning

September 29, 2023

Adversarial Driving Behavior Generation Incorporating Human Risk Cognition for Autonomous Vehicle Evaluation
Zhen Liu, Hang Gao, Hao Ma, Shuo Cai, Yunfeng Hu, Ting Qu, Hong Chen, Xun Gong
Autonomous Vehicle Deep Deterministic Policy Gradient Adversarial Behavior Adversarial Policy Autonomous System Testing

August 22, 2023

Robust Lagrangian and Adversarial Policy Gradient for Robust Constrained Markov Decision Processes
David M. Bossens
Reinforcement Learning Policy Gradient Robust Policy Lagrangian System Adversarial Reinforcement Learning Adversarial Policy

May 27, 2023

Rethinking Adversarial Policies: A Generalized Attack Formulation and Provable Defense in RL
Xiangyu Liu, Souradip Chakraborty, Yanchao Sun, Furong Huang
Adversarial Attack Human Like RL Adversarial Policy Step Attack Provable Defense

May 4, 2023

Toward Evaluating Robustness of Reinforcement Learning with Adversarial Policy
Xiang Zheng, Xingjun Ma, Shengjie Wang, Xinyu Wang, Chao Shen, Cong Wang
Reinforcement Learning Robustness Evaluation Adversarial Agent Adversarial Policy State Adversarial

April 27, 2023

Adversarial Policy Optimization in Deep Reinforcement Learning
Md Masudur Rahman, Yexiang Xue
Deep Reinforcement Learning Policy OpTimization Based Policy Adversarial Policy PPO Algorithm Deep Reinforcement Learning Algorithm

February 27, 2023

Implicit Poisoning Attacks in Two-Agent Reinforcement Learning: Adversarial Policies for Training-Time Attacks
Mohammad Mohammadi, Jonathan Nöther, Debmalya Mandal, Adish Singla, Goran Radanovic
Poisoning Attack Training Time Attack Adversarial Policy Optimal Adversarial Optimal Attack

November 1, 2022

Adversarial Policies Beat Superhuman Go AIs
Tony T. Wang, Adam Gleave, Tom Tseng, Kellin Pelrine, Nora Belrose, Joseph Miller, Michael D. Dennis, Yawen Duan, Viktor Pogrebniak, Sergey Levine, Stuart Russell
Adversarial Policy Superhuman Artificial Intelligence

October 30, 2022

Imitating Opponent to Win: Adversarial Policy Imitation Learning in Two-player Competitive Games
The Viet Bui, Tien Mai, Thanh H. Nguyen
Imitation Learning Two Player Adversarial Policy

September 5, 2022

Red Teaming with Mind Reading: White-Box Adversarial Policies Against RL Agents
Stephen Casper, Taylor Killian, Gabriel Kreiman, Dylan Hadfield-Menell
Adversarial Example Reinforcement Learning Agent Red Teaming Adversarial Agent Adversarial Policy White Box Adversarial

August 10, 2022

Reducing Exploitability with Population Based Training
Pavel Czempin, Adam Gleave
Adversarial Training Zero Sum Game Adversarial Policy Monotone Adversary Population Based Training Self Play Reinforcement Learning

June 10, 2022

Adversarial Counterfactual Environment Model Learning
Xiong-Hui Chen, Yang Yu, Zheng-Mao Zhu, Zhihua Yu, Zhenjun Chen, Chenghe Wang, Yinan Wu, Hongqiu Wu, Rong-Jun Qin, Ruijin Ding, Fangsheng Huang
Policy Learning Adversarial Policy Counterfactual Generative

April 9, 2022

Improve Generalization of Driving Policy at Signalized Intersections with Adversarial Learning
Yangang Ren, Guojian Zhan, Liye Tang, Shengbo Eben Li, Jianhua Jiang, Jingliang Duan
Autonomous Driving Strong Generalization Autonomous Vehicle Adversarial Learning Adversarial Policy Driving Policy Signalized Intersection Dynamic Traffic

Adversarial Policy

Papers

ResTNet: Defense against Adversarial Policies via Transformer in Computer Go

VCAT: Vulnerability-aware and Curiosity-driven Adversarial Training for Enhancing Autonomous Vehicle Robustness

FREA: Feasibility-Guided Generation of Safety-Critical Scenarios with Reasonable Adversariality

Overcoming Reward Overoptimization via Adversarial Policy Optimization with Lightweight Uncertainty Estimation

SUB-PLAY: Adversarial Policies against Partially Observed Multi-Agent Reinforcement Learning Systems

From Principle to Practice: Vertical Data Minimization for Machine Learning

Posterior Sampling for Competitive RL: Function Approximation and Partial Observation

Quantifying Assistive Robustness Via the Natural-Adversarial Frontier

Adversarial Driving Behavior Generation Incorporating Human Risk Cognition for Autonomous Vehicle Evaluation

Robust Lagrangian and Adversarial Policy Gradient for Robust Constrained Markov Decision Processes

Rethinking Adversarial Policies: A Generalized Attack Formulation and Provable Defense in RL

Toward Evaluating Robustness of Reinforcement Learning with Adversarial Policy

Adversarial Policy Optimization in Deep Reinforcement Learning

Implicit Poisoning Attacks in Two-Agent Reinforcement Learning: Adversarial Policies for Training-Time Attacks

Adversarial Policies Beat Superhuman Go AIs

Imitating Opponent to Win: Adversarial Policy Imitation Learning in Two-player Competitive Games

Red Teaming with Mind Reading: White-Box Adversarial Policies Against RL Agents

Reducing Exploitability with Population Based Training

Adversarial Counterfactual Environment Model Learning

Improve Generalization of Driving Policy at Signalized Intersections with Adversarial Learning