Rewriting modulo isotopies in Khovanov-Lauda-Rouquier's categorification of quantum groups (Q2219329)

The main objective of this paper is to show that a given 2-category is a categorification of some quantum group using high-order rewriting techniques. It has three sections: Section 1 concerns the presentation of notions of linear (3,2)-polygraphs used to provide presentation by generators and relations for linear 2-categories, Section 2 provides a convergent presentation of the simply-laced KLR algebras and finally Section 3 exploits techniques of rewriting modulo to show that Khovanov-Lauda-Rouquier's 2-category (in a particular case) is indeed a categorification of a given quantum group. 2-Categorification of algebras roughly means the process of replacing an algebra (and its actions on vector spaces) by, for instance, an additive 2-category (and functorial actions) whose Grothendieck group is isomorphic to the algebra under consideration. The paper is devoted to a detailed analysis of Khovanov and Lauda's 2-category \(\mathcal{U}(\mathfrak{g})\) (see [\textit{M. Khovanov} and \textit{A. D. Lauda}, Quantum Topol. 1, No. 1, 1--92 (2010; Zbl 1206.17015)]) associated to a symmetrizable Kac-Moodly algebra \(\mathfrak{g}\), considered as a categorification of a quantum group introduced by \textit{G. Lusztig} [Introduction to quantum groups. Reprint of the 1994 ed. Boston, MA: Birkhäuser (2010; Zbl 1246.17018)]. More precisely two distinct linear 2-categories are studied using rewriting techniques, namely linear (3,2)-polygraphs as introduced in [\textit{C. Alleaume}, Higher-dimensional linear rewriting and coherence in categorification and representation theory, Université de Lyon, June 2018] and [\textit{B. Dupont}, ``Rewriting modulo isotopies in pivotal linear (2,2)-categories'', preprint, \url{arXiv:1906.03904}]. Firstly (in Section 2), the Khovanov-Lauda-Rouquier (KLR) algebras (see [\textit{M. Khovanov} and \textit{A. D. Lauda}, Represent. Theory 13, 309--347 (2009; Zbl 1188.81117)]) are introduced (Definition 2.2.1, p.15). They are particular instances of more general KLR algebras, namely they are unsigned versions thereof. The author shows that the KLR algebra thus introduced is encoded in a linear 2-category \(\mathcal{C}^{KLR}\) under the direct sum of its spaces of 2-cells. The author introduces a linear (3,2)-polygraph KLR which provides a convergent presentation of the linear 2-category \(\mathcal{C}^{KLR}\) (Theorem 2.3.2, p.19). It is also shown (Corollary 2.3.6, p.24) that the simply-laced KLR algebras admit Poincaré-Birkhoff-Witt bases in the sense of [\textit{R. Rouquier}, ``2-Kac-Moody algebras'', Preprint, \url{arXiv:0812.5023}]. Secondly (Section 3), the author studies the Khovanov-Lauda-Rouquier's 2-categorification \(\mathcal{U}(\mathfrak{g})\) of a certain quantum group, for a simply-laced symmetrizable Kac-Moody algebra \(\mathfrak{g}\), using techniques of linear (3,2)-polygraphs modulo (that is, rewriting modulo some non-oriented axioms). The author introduces a polygraphic presentation \(\mathcal{KLR}\) of \(\mathcal{U}(\mathfrak{g})\) and with Theorem 3.5.3, provides an explicit description of a linear basis of the spaces of 2-cells of \(\mathcal{U}(\mathfrak{g})\). Consequently this shows that \(\mathcal{U}(\mathfrak{g})\) is a 2-categorification of a certain quantum group associated with \(\mathfrak{g}\) (this also uses results from [\textit{M. Khovanov} and \textit{A. D. Lauda}, Quantum Topol. 1, No. 1, 1--92 (2010; Zbl 1206.17015)]).