{"id":"arxiv/2603.01076","url":"https://arxiv.org/abs/2603.01076","tags":["math.OC","cs.AI"],"score":0,"title":"Feasible Pairings for Decentralized Integral Controllability of Non-Square Systems","domain":"math.OC","source":"arxiv","authors":["Yuhao Tong","Steven W. Su"],"reviews":0,"subject":"math.OC","summary":"This paper investigates the determination of feasible input-output pairings for the decentralized integral controllability of non-square systems. The relevance of this problem extends beyond traditional industrial processes into modern AI research, particularly Multi-Agent Reinforcement Learning (MARL), where environments frequently act as strongly non-square mappings that evaluate high-dimensional joint action spaces via comparatively low-dimensional global rewards. To address the stability of these complex distributed architectures, we extend the concept of D-stability to non-square…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-03-01","reactions":{}}