An introduction to representation-theoretic canonical bases of cluster algebras
Fan Qin
TL;DR
This paper develops a representation-theoretic lens on cluster algebras by introducing and analyzing the common triangular (Kazhdan-Lusztig-type) bases $\mathbf{L}$, connecting them to dual canonical bases $\mathbf{B}^*$ in Lie-theoretic contexts. It builds seeds from signed words, details cluster operations (freezing, base change), and proves how the KL algorithm and categorification ideas produce robust bases that are compatible across mutation sequences and extend to infinite ranks via colimits. The results demonstrate the existence and propagation of common triangular bases for many cluster algebras arising from Lie theory, including quasi-categorifications in symmetric cases, and provide a framework to transfer structures through extensions and reductions of seeds. The work equips researchers with concrete tools to study canonical bases in cluster algebras, with explicit constructions and examples from quantum groups, unipotent cells, and quantum affine algebras, and offers an overarching perspective on extending these bases to infinite-dimensional settings.
Abstract
This is an introduction to cluster algebras and their common triangular bases. These bases are Kazhdan-Lusztig-type and serve as the canonical bases of cluster algebras from the representation-theoretic point of view. We review seeds associated with signed words, cluster operations (freezing and base change), the extension technique from finite to infinite rank, which are convenient tools to study these algebras and their bases. We present recent results on the topic and provide examples. An appendix collects detailed examples of some cluster algebras from Lie theory, including those from Lie groups and representations of quantum affine algebras.
