Journal Article10.48550/arXiv.2306.06136
Robustness Testing for Multi-Agent Reinforcement Learning: State Perturbations on Critical Agents
Ziyuan Zhou,Guanjun Liu +1 more
TL;DR: In this paper , the authors proposed a robustness testing framework for multi-agent reinforcement learning (MARL) that attacks states of Critical Agents (RTCA), which has two innovations: a Differential Evolution (DE) based method to select critical agents as victims and to advise the worst-case joint actions on them; and a team cooperation policy evaluation method employed as the objective function for the optimization of DE.
read more
Abstract: Multi-Agent Reinforcement Learning (MARL) has been widely applied in many fields such as smart traffic and unmanned aerial vehicles. However, most MARL algorithms are vulnerable to adversarial perturbations on agent states. Robustness testing for a trained model is an essential step for confirming the trustworthiness of the model against unexpected perturbations. This work proposes a novel Robustness Testing framework for MARL that attacks states of Critical Agents (RTCA). The RTCA has two innovations: 1) a Differential Evolution (DE) based method to select critical agents as victims and to advise the worst-case joint actions on them; and 2) a team cooperation policy evaluation method employed as the objective function for the optimization of DE. Then, adversarial state perturbations of the critical agents are generated based on the worst-case joint actions. This is the first robustness testing framework with varying victim agents. RTCA demonstrates outstanding performance in terms of the number of victim agents and destroying cooperation policies.
read more
Chat with Paper
AI Agents for this Paper
Find similar papers on Google Scholar, PubMed and Arxiv
Write a critical review of this paper
Analyze citations of this paper to find unaddressed research gaps
Citations
Robust Multi-Agent Reinforcement Learning Method Based on Adversarial Domain Randomization for Real-World Dual-UAV Cooperation
TL;DR: An adversarial domain randomization method that utilizes an adversarial generator as a “nature player” to generate a more reasonable training environment so that the trained decision policy can deal with complex situations is proposed.
12
A simulation and experimentation architecture for resilient cooperative multiagent reinforcement learning models operating in contested and dynamic environments
Ishan Honhaga,Claudia Szabo +1 more
TL;DR: A simulation and experimentation architecture for resilient cooperative multiagent reinforcement learning models operating in contested and dynamic environments explores the resilience of c-MARL models against various attacks and the impact of training with different perturbations on their effectiveness.
2
Probabilistic Automata-Based Method for Enhancing Performance of Deep Reinforcement Learning Systems
Min Yang,Guanjun Liu,Ziyuan Zhou,Jiacun Wang +3 more
TL;DR: A probabilistic automata-based framework enhances deep reinforcement learning (DRL) performance by proactively identifying and correcting vulnerabilities through real-time monitoring and correction mechanisms, improving model performance with minimal modifications in industrial manufacturing domains.
1
Byzantine Robust Cooperative Multi-Agent Reinforcement Learning as a Bayesian Game
TL;DR: In this paper , a Bayesian Adversarial Robust Dec-POMDP framework is proposed to address the uncertainty that any agent can be adversarial, which views Byzantine adversaries as nature-dictated types, represented by a separate transition.
Measuring the Robustness of Multi-Agent Reinforcement Learning Systems under Partial Agent Failure
Zoltán Barta,Balázs Nagy,László Gulyás +2 more
- 23 Jun 2025
TL;DR: This study examines the robustness of multi-agent reinforcement learning systems under partial agent failure, revealing that reward function structure and neural network generalization properties significantly impact system resilience, with design lessons for real-world deployments.
References
Differential Evolution – A Simple and Efficient Heuristic for Global Optimization over Continuous Spaces
Rainer Storn,Kenneth Price +1 more
TL;DR: In this article, a new heuristic approach for minimizing possibly nonlinear and non-differentiable continuous space functions is presented, which requires few control variables, is robust, easy to use, and lends itself very well to parallel computation.
•Posted Content
Proximal Policy Optimization Algorithms
TL;DR: A new family of policy gradient methods for reinforcement learning, which alternate between sampling data through interaction with the environment, and optimizing a "surrogate" objective function using stochastic gradient ascent, are proposed.
18K
•Posted Content
Explaining and Harnessing Adversarial Examples
TL;DR: The authors argue that the primary cause of neural networks' vulnerability to adversarial perturbation is their linear nature, which is supported by new quantitative results while giving the first explanation of the most intriguing fact about adversarial examples: their generalization across architectures and training sets.
15.9K
•Posted Content
Towards Deep Learning Models Resistant to Adversarial Attacks
TL;DR: This work studies the adversarial robustness of neural networks through the lens of robust optimization, and suggests the notion of security against a first-order adversary as a natural and broad security guarantee.
•Posted Content
Multi-Agent Actor-Critic for Mixed Cooperative-Competitive Environments
TL;DR: An adaptation of actor-critic methods that considers action policies of other agents and is able to successfully learn policies that require complex multi-agent coordination is presented.
2.9K