Backend
Constrained Black-Box Attacks Against Cooperative Multi-Agent Reinforcement Learning
arXiv cs.MA0 views
arXiv:2508.09275v3 Announce Type: replace-cross Abstract: Collaborative multi-agent reinforcement learning (MARL) is increasingly deployed in critical domains, yet its vulnerability to adversarial perturbations has not been thoroughly investigated. Most existing adversarial attacks against MARL rely on optimistic assumptions, such as access to policy weights or the ability to train surrogate policies. We investigate whether adversaries with limited capabilities remain effective against MARL agents. In particular, we study two restricted attack settings that are overlooked in the literature: (i) adversaries with access only to agent observations and (ii) adversaries with no access whatsoever. Our approach generates perturbations that intentionally misalign how victim agents perceive their environment, preventing them from collaborating. We conduct extensive experiments across a diverse set of 22 environments. Our results demonstrate both the effectiveness and sample efficiency of our attacks.
Source：https://arxiv.org/abs/2508.09275PreviousWhen Is Fresh AI Stock Analysis Worth It? Allocating a Limited Research Budget Across Price Updates, Deep Forecast Refreshes, and Editorial Reviews of Top Stock PicksNextMulti-User Large Language Model Agents
Related posts
Quantum estimates for classical polynomial optimization
0arXiv cs.NA
A Tensor-Network Chebyshev--Galerkin Method for Boundary-Value Problems in Computational Mechanics
0arXiv cs.NA
Linear adaptive variable time step schemes for the anisotropic phase field dendritic crystal growth model
0arXiv cs.NA
Magnetic Bloch eigenvalue problems in nonuniform periodic magnetic fields: a Fourier-finite difference approach
0arXiv cs.NA
Phantoms and Disclosures: A Statistical Framework for Auditing Privacy in Synthetic Data
0arXiv stat.ML
When the Agent Learns Too: Learning to Incentivize with Unobserved Rewards
0arXiv stat.ML