Imaginary modules arising from tensor products of snake modules
Matheus Brito, Adriano Moura · 2025
Read Imaginary modules arising from tensor products of snake modules by Matheus Brito, Adriano Moura (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-21159.txt.
Introduction
Matheus Brito and Adriano Moura study special building blocks in the representation theory of quantum loop algebras. Their paper from twenty twenty-five deals with modules, structured spaces on which the algebra acts. Some modules are called imaginary, and cluster algebra tools miss them. The authors widen a frame built with Chari to find more of them.
Their work starts from tensor products of snake modules. The first main result describes the socle, the simple base layer, of such products when the snakes form a covering pair. A broader covering link yields a chain of nested submodules. The authors guess that each step's quotient is simple and imaginary, save the socle. They prove the first quotient is simple, and imaginary under a mild extra term. New classes of imaginary modules result.
Catalog record
| CREATOR | Matheus Brito, Adriano Moura |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 177 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2505-21159
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Imaginary modules arising from tensor products of snake modules Thanks: The work of M.B. and A.M. was partially supported CNPq grant 405793/2023-5. The work of A.M. was also partially supported by Fapesp grant 2018/23690-6. Matheus Brito, Adriano Moura Abstract . Motivated by the limitations of cluster algebra techniques in detecting imaginary modules, we build on the representation-theoretic framework developed by the first author and Chari to extend the construction of such modules beyond previously known cases, which arise from the tensor product of a higher-order Kirillov–Reshetikhin module and its dual. Our first main result gives an explicit description of the socle of tensor products of two snake modules, assuming the corresponding snakes form a covering pair of ladders. By considering a higher-order generalization of the covering relation, we describe a sequence of inclusions of highest- ℓ \ell -weight submodules of such tensor products. We conjecture all the quotients of subsequent modules in this chain of inclusions are simple and imaginary, except for the socle itself, which might be real. We prove the first such quotient is indeed simple and, assuming an extra mild condition, we also prove it is imaginary, thus giving rise to new classes of imaginary modules within the category of finite-dimensional representations of quantum loop algebras in type A A . 1. Introduction This paper is dedicated to the study of certain aspects of the tensor product of two prime snake modules for a quantum loop algebra U_{q}(\tilde{\mathfrak{g}}) in the case that the finite-dimensional simple Lie algebra \mathfrak{g} is of type A . We assume q is neither zero nor a root of unity. Since the discovery by Hernandez-Leclerc [ 12 , 13 ] that the Grothendieck rings of certain proper subcategories of the abelian category of finite-dimensional U_{q}(\tilde{\mathfrak{g}}) -modules are cluster algebras, the task o classifying the prime and the real simple objects of the category became one of the most studied topics in the area. Since a module is real if its tensor square is simple, the task of classifying the real modules is equivalent to that of classifying the simple modules which are not real. Such modules were termed imaginary in [ 19 ] , where the first example of an imaginary simple module was given. The results of [ 1 , 16 , 24 , 26 ] imply that, in principle, the real modules can be determined using cluster algebra machinery. In practice, however, determining if a given Drinfeld polynomial arises from a cluster monomial is not a simple task, which motivates the search for alternative techniques. For instance, this is the spirit of [ 3 ] . Furthermore, the combinatorial framework of cluster algebras does not provide representation theoretic insights for the existence of imaginary modules. Seeking for such insights is the spirit of the present paper. The class of snake modules introduced in [ 23 ] has been proven to be particularly important in many aspects. In particular, their q -characters can be computed by a relatively simple combinatorial data, which we refer to as Mukhin-Young (MY) paths, and the connection with cluster monomials can be made explicitly, thus showing snake modules are real. The class of Kirillov-Reshetikhin (KR) modules and, more generally, that of minimal affinizations, are proper subclasses of that of (prime) snake modules. In [ 2 ] , the authors considered the subclass of prime snake modules whose highest weight is a multiple of a fundamental weight, thus terming such modules “higher order KR modules”. Suppose V is such a module. By studying V\otimes V^{*} , a new family of imaginary modules was described in [ 2 ] , assuming N={\rm rank}(\mathfrak{g})>2 . The pioneering example of imaginary modules from [ 19 ] is recovered by this construction. The main results of the present paper substantially extends this construction of imaginary modules by studying much more general tensor products of two prime snake modules. The assumption N>2 is implied by the assumptions of the construction. We remark that [ 2 ] also gives a different construction specific to the case N=2 (and it is well known that all simple modules are real if N=1 ). Before continuing, let us recall that the isomorphism classes of simple modules of the category of (type 1) finite-dimensional U_{q}(\tilde{\mathfrak{g}}) -modules are classified in terms of Drinfeld polynomials. We denote by \mathcal{P}^{+} the set of all Drinfeld polynomials, which is a multiplicative monoid in a natural way. For \boldsymbol{\pi}\in\mathcal{P}^{+} , we let V(\boldsymbol{\pi}) be any element in the corresponding isomorphism class. The trivial Drinfeld polynomial will be denoted by \boldsymbol{1} . As a “preliminary” step towards our construction of imaginary modules, the first of our main results, Theorem 2.5.1 , describes explicitly the socle of the tensor products of two prime snake modules, assuming the corresponding snakes form what we termed a covering pair of ladders. Let us explain what this means. Without loss of generality, we restrain ourselves to the subcategory consisting of modules which have simple factors whose Drinfeld polynomials have roots in q^{\mathbb{Z}} . Let \tilde{\mathbb{S}} be a parameterizing set for the isomorphism classes of the corresponding fundamental modules and let \mathbb{S}=\tilde{\mathbb{S}}\cup\{\emptyset\} . In the main part of the text, we chose to work with \tilde{\mathbb{S}} being the set of certain “segments”. For s\in\tilde{\mathbb{S}} , let \boldsymbol{\omega}_{s} denote the corresponding Drinfeld polynomial, while \boldsymbol{\omega}_{\emptyset}=\boldsymbol{1} . Given s,s^{\prime}\in\tilde{\mathbb{S}} , we say s covers s^{\prime} , and write s\rhd s^{\prime} , if (1) V(\boldsymbol{\omega}_{s})\otimes V(\boldsymbol{\omega}_{s^{\prime}})\ \ \text{is reducible and highest-$\ell$-weight.} Recall that a module V is highest- \ell -weight if its set of weights contains a unique maximal element, the corresponding weight space is one-dimensional, and V is cyclically generated by this weight space. In the language of segments, ( 1 ) can be phrased in a completely intrinsic manner, with no mention to representation theory (see Section 2.4 ). It is well-known that, if ( 1 ) is satisfied, then V(\boldsymbol{\omega}_{s})\otimes V(\boldsymbol{\omega}_{s^{\prime}}) has length 2 and its socle is a simple module of the form V(\boldsymbol{\omega}_{s\cap s^{\prime}})\otimes V(\boldsymbol{\omega}_{s\cup s^{\prime}})\cong V(\boldsymbol{\omega}_{s\diamond s^{\prime}})\ \ \text{where}\ \ \boldsymbol{\omega}_{s\diamond s^{\prime}}=\boldsymbol{\omega}_{s\cap s^{\prime}}\,\boldsymbol{\omega}_{s\cup s^{\prime}} and s\cap s^{\prime},s\cup s^{\prime}\in\mathbb{S} . In the language of segments, s\cap s^{\prime} is indeed the intersection of the corresponding segments and similarly for s\cup s^{\prime} , thus the choice of notation. A prime snake of length l corresponds to a sequence \boldsymbol{s}=s_{1},s_{2},\dots,s_{l} of elements in \tilde{\mathbb{S}} such that s_{k+1}\rhd s_{k}\ \ \text{for all}\ \ 1\leq k<l. Such sequences are also called ladders. In that case, V(\boldsymbol{\omega}_{s_{l}})\otimes\cdots\otimes V(\boldsymbol{\omega}_{s_{1}}) is highest- \ell -weight and the associated prime snake module is the simple quotient of this tensor product. Setting \boldsymbol{\omega}_{\boldsymbol{s}}=\prod_{k=1}^{l}\boldsymbol{\omega}_{s_{k}}, it follows that this simple quotient is isomorphic to V(\boldsymbol{\omega}_{\boldsymbol{s}}) . If \boldsymbol{s},\boldsymbol{s}^{\prime} are ladders of length l , we say \boldsymbol{s} covers \boldsymbol{s}^{\prime} and write \boldsymbol{s}\rhd\boldsymbol{s}^{\prime} if s_{k}\rhd s^{\prime}_{k}\ \ \text{for all}\ \ 1\leq k\leq l. Theorem 2.5.1 says that \boldsymbol{s}\rhd\boldsymbol{s}^{\prime}\quad\Rightarrow\quad{\rm soc}(V(\boldsymbol{\omega}_{\boldsymbol{s}})\otimes V(\boldsymbol{\omega}_{\boldsymbol{s}^{\prime}}))\cong V(\boldsymbol{\omega}_{\boldsymbol{s}\diamond\boldsymbol{s}^{\prime}})\quad\text{where}\quad\boldsymbol{\omega}_{\boldsymbol{s}\diamond\boldsymbol{s}^{\prime}}=\prod_{k=1}^{l}\boldsymbol{\omega}_{s_{k}\diamond s^{\prime}_{k}}. A closely related result was obtained in [ 17 , Corolary 6.16] in the framework of complex smooth representations of GL_{n}(\mathbb{F}) with \mathbb{F} being an nonarchimedian field. Using a Schur-Weyl type duality [ 5 ] (we refer the reader to [ 9 ] for a deeper discussion), this result gives rise to an algorithm for computing the socle of a