Constructing Chayet–Garibaldi algebras from affine vertex algebras (including the 3876 3876 -dimensional algebra for E 8 E_{8} )

Tom De Medts, Louis Olyslager · 2025

Read Constructing Chayet–Garibaldi algebras from affine vertex algebras (including the 3876 3876 -dimensional algebra for E 8 E_{8} ) by Tom De Medts, Louis Olyslager (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-2503-24103.txt.

Introduction

Tom De Medts and Louis Olyslager wrote a 2025 algebra paper that reframes a recent family of algebras. In 2021, Chayet and Garibaldi built a commutative, non-associative algebra for each simple Lie algebra. Their formulas were explicit but ad hoc. One case sits on a huge E8 representation with a unit added.

De Medts and Olyslager found a natural home for these algebras in affine vertex algebras. In that language, the ad hoc formulas turn strikingly simple. The new view may explain why the construction works at all. It could also guide the study of this algebra class. The paper links two fields that rarely meet.

Catalog record

CREATORTom De Medts, Louis Olyslager
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE55 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2503-24103

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Constructing Chayet–Garibaldi algebras from affine vertex algebras (including the 3876 3876 -dimensional algebra for E 8 E_{8} ) Tom De Medts, Louis Olyslager Abstract In 2021, Maurice Chayet and Skip Garibaldi provided an explicit construction of a commutative non-associative algebra on the second smallest representation of E 8 E_{8} (of dimension 3875 3875 ) adjoined with a unit. In fact, they define such an algebra A ⁡ ( 𝔤 ) A(\mathfrak{g}) for each simple Lie algebra 𝔤 \mathfrak{g} , in terms of explicit but ad-hoc formulas. We discovered that their algebras A ⁡ ( 𝔤 ) A(\mathfrak{g}) have a natural interpretation in terms of affine vertex algebras, and their ad-hoc formulas take an extremely simple form in this new interpretation. It is our hope that this point of view will lead to a better understanding of this interesting class of algebras. MSC2020. 20G41, 17B45, 17B69, 20G05, 20G15 1 Introduction The second smallest irreducible representation of algebraic groups of type E_{8} , which is of dimension 3875 , has received some attention lately, because it admits a (unique) invariant commutative non-associative algebra product. This fact itself is not surprising, but it was not until a few years ago that this algebra has been constructed explicitly. This had been achieved almost simultaneously by the first author and Michiel Van Couwenberghe [ 3 ] and by Maurice Chayet and Skip Garibaldi [ 2 ] . In fact, in both constructions, the algebra is a 3876 -dimensional algebra obtained by adjoining a unit, and the construction relies on the symmetric square of the Lie algebra. The fact that the resulting algebra has as its automorphism group precisely the original group of type E_{8} (and nothing more) is a consequence of [ 6 ] . The construction of Chayet and Garibaldi from [ 2 ] is quite general. In fact, it takes as input • any Lie algebra \mathfrak{g} over any field k with \Char(k)\neq 2 , • a non-degenerate invariant symmetric bilinear form \kappa on \mathfrak{g} , and produces a commutative non-associative algebra that we will denote by A(\mathfrak{g},\kappa) . Chayet and Garibaldi only considered the case where \mathfrak{g} is a simple Lie algebra, in which case \kappa is necessarily a non-zero multiple of the Killing form K , and they only use the fixed scaling \kappa=\frac{1}{2h^{\vee}}K , where h^{\vee} is the dual Coxeter number of \mathfrak{g} . Their algebra is simply denoted by A(\mathfrak{g}) . The explicit formulas in the construction of Chayet and Garibaldi are very much ad hoc —see Section 2 below—and were discovered by “trial and error”. The construction in [ 3 ] is slightly different and builds the algebra piece by piece, and therefore gives perhaps even less insight in the algebraic structure. It was an unexpected and pleasant surprise to us that we recognized the formulas of [ 2 ] in a quite unrelated area, namely in the theory of vertex (operator) algebras . More precisely, we will be able to construct the algebras A(\mathfrak{g},\kappa) as substructures of a simple vertex algebra L_{\widehat{\mathfrak{g}}}(1,0) arising from 1 1 1 We point out that it is common to define vertex algebras L_{\widehat{\mathfrak{g}}}(\ell,0) “of level \ell ” for any constant \ell ; see, for example, [ 10 , §6.2] . We have decided to incorporate this constant in the construction of \widehat{\mathfrak{g}} instead, because the resulting formulas are easier while we do not lose any generality. (Notice, in particular, that we do not normalize the bilinear form as in [ 10 , Remark 6.2.15] .) See also Remark 4.4 below. the affine Lie algebra \widehat{\mathfrak{g}} corresponding to \mathfrak{g} . Vertex algebras have been introduced by Richard Borcherds in 1986 (see [ 1 ] ) and have famously played a key role in the monstrous moonshine. We follow Borcherds’ original approach and we allow for an arbitrary base field k with \Char(k)\neq 2 . For the concrete setup, we will essentially follow [ 10 ] , but modified to allow for such arbitrary fields k , as in [ 11 ] . Usually, vertex algebras are defined using formal Laurent series, but we prefer to avoid this setup and work directly with the infinitely many bilinear “multiplication maps”, for two reasons. First, we will need to explicitly compute some of these multiplication maps anyway. Second, this also helps to make our paper more accessible to people who are not at all familiar with the theory of vertex algebras. Our main result can be summarized as follows. (We refer to the text for a detailed explanation of the notation used in this statement.) Main Theorem . Let \mathfrak{g} be a Lie algebra over a field k with \Char(k)\neq 2 with Z(\mathfrak{g})=0 , equipped with a non-degenerate invariant symmetric bilinear form \kappa . Let V be the \mathbb{N} -graded affine vertex algebra V=V_{\widehat{\mathfrak{g}}}(1,0) (which depends also on \kappa ) and let L be the quotient of V by its unique maximal graded ideal, which is a simple graded vertex algebra. Let L_{(2)}^{\mathrm{sym}} be the subspace of L_{(2)} spanned by the elements a_{-1}b+b_{-1}a for all a,b\in L_{(1)} . Make L_{(2)}^{\mathrm{sym}} into a commutative algebra by the “Jordan product” a\bullet b\coloneqq\tfrac{1}{2}(a_{1}b+b_{1}a) and define a bilinear form on L_{(2)}^{\mathrm{sym}} by (a,b)\mapsto a_{3}b for all a,b\in L_{(2)}^{\mathrm{sym}} . Then L_{(2)}^{\mathrm{sym}} is isomorphic to the Chayet–Garibaldi algebra A(\mathfrak{g},\kappa) (as an algebra with bilinear form, up to scaling of the bilinear form). It is our hope that this construction will provide new insight into the Chayet–Garibaldi algebras A(\mathfrak{g},\kappa) , including the notorious (3875+1) -dimensional algebra for E_{8} . Outline of the paper In Section 2 , we review the construction of the class of algebras A(\mathfrak{g}) constructed by Chayet and Garibaldi, generalized to arbitrary Lie algebras equipped with a non-degenerate symmetric bilinear invariant form. In Section 3 , we give the required background on vertex algebras (in arbitrary characteristic \neq 2 ). We have, on purpose, restricted this to the bare minimum while keeping the paper self-contained. In the next Section 4 , we then present the construction of affine vertex algebras (i.e., vertex algebras arising from affine Lie algebras), translated to our setup from Section 3 . At the end of this section, we collect a number of useful identities. Section 5 , finally, is the core of the paper. We combine the Chayet–Garibaldi algebras from Section 2 with the affine vertex algebras from Section 4 , culminating in our main result in Theorem 5.24 . 2 The Chayet–Garibaldi algebras A(\mathfrak{g},\kappa) We start by summarizing the construction of the algebras A(\mathfrak{g},\kappa) introduced by Chayet and Garibaldi in [ 2 ] . We give their construction in a more general setting than they do, by allowing for arbitrary Lie algebras equipped with a non-degenerate symmetric bilinear invariant form. (See 2.3 below for the precise setting that we use.) Definition 2.1 . Let k be a commutative field, let \mathfrak{g} be a Lie algebra over k and let \kappa\colon\mathfrak{g}\times\mathfrak{g}\to k\colon(a,b)\mapsto\langle a,b\rangle be a symmetric bilinear form on \mathfrak{g} . 1. We call \kappa (Lie or \mathfrak{g} -) invariant if \langle[a,b],c\rangle=\langle a,[b,c]\rangle for all a,b,c\in\mathfrak{g} . 2. We call \kappa non-degenerate if the only a\in\mathfrak{g} for which \langle a,\mathfrak{g}\rangle=0 is a=0 . Remark 2.2 . For a finite-dimensional simple Lie algebra, the

Read the complete text