📝 SummarXiv

Abstract

Recent papers in solvable lattice models emphasize models where states can be visualized as colored paths through the lattice. We define a bosonic model in which there are two types of colors, one whose paths move down and to the right, the other whose paths move down and to the left. Depending on their boundary data, systems may have no states, exactly one state, or many states. We prove that these cases depend on a criterion involving two permutations extracted from the boundary data and their Bruhat order. This classification also helps us to characterize the partition functions of our systems, a question at the heart of the study of solvable lattice models. Using the solvability of the model, we derive a four-term recurrence relation on the partition function. Together with the classification of systems by number of states which serves as a base case for the recursion, the recursion completely characterizes the partition function of systems. We also show a color merging property relating the bicolored bosonic models to colored and uncolored bosonic models, and correspondence with Gelfand-Tsetlin patterns.