{"id":"arxiv/2603.18304","url":"https://arxiv.org/abs/2603.18304","tags":["math.OC"],"score":0,"title":"Forward-Backward Dynamic Programming for LQG Dynamic Games with Partial and Asymmetric Information","domain":"math.OC","source":"arxiv","authors":["Yuxiang Guan","Iman Shames","Tyler Summers"],"reviews":0,"subject":"math.OC","summary":"We formulate and study a class of two-player zero-sum stochastic dynamic games with partial and asymmetric information. Information asymmetry introduces fundamental challenges involving \\emph{belief representation} and \\emph{theory of mind} issues, where agents must impute belief states and estimates of other agents to inform their own strategy. To avoid an infinite regress of higher-order beliefs amongst agents and obtain computationally implementable results, we focus on a linear quadratic Gaussian (LQG) model and consider strategies with limited internal state dimension. We present a…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-03-18","reactions":{}}