Id: arxiv/2611.22557 Url: https://arxiv.org/abs/2611.22557 Tags: math.OC, cs.LG Score: 0 Title: Balancing Exploration and Exploitation in Constrained Combinatorial Optimization with U(1)-Symmetric Tensor Trains Domain: math.OC Source: arxiv Authors: Sergei Iudin, Mikhail Podobrii, Dmitry Zheltkov, Stanislav Moiseev Reviews: 0 Subject: math.OC Category: research Comments: 0 Disputed: false Division: math Mentions: 0 Discussed: 0 Published: 2026-11-17 Page: https://delphi.science/r/arxiv/2611.22557 Tensor Train (TT) decomposition has emerged as a powerful tool for mitigating the curse of dimensionality in a variety of machine learning and combinatorial optimization tasks. Recent advances show that TT-based generative models can efficiently represent and explore high-dimensional discrete spaces, even when the cost function structure is unknown. However, incorporating hard linear constraints into TT-based optimization remains challenging, as existing interpolation and sampling approaches often fail when feasible points are rare. In this work, we develop a general TT-based framework for…