Papers
Topics
Authors
Recent
Search
2000 character limit reached

Kadar–Yu Algebras: A Diagrammatic Framework

Updated 2 January 2026
  • Kadar–Yu algebras are a family of diagrammatic algebras that interpolate between Temperley–Lieb and Brauer algebras by imposing a left-height constraint on Brauer diagrams.
  • They are generated by Temperley–Lieb and symmetric group elements, using diagrammatic multiplications and mixed relations to capture rich representation-theoretic structures.
  • Their structure features iterated inflation, Chebyshev-type recurrence relations, and ties to Kazhdan–Lusztig theory and alcove geometry, unifying semisimple and non-semisimple phenomena.

The Kadar–Yu (KY) algebras form a distinguished tower of diagrammatic algebras interpolating between the Temperley–Lieb and Brauer algebras. Parameterized by a non-negative integer \ell and a parameter α\alpha or δ\delta from the base field KK (typically K=CK = \mathbb{C} or an algebraically closed field of characteristic zero), these algebras arise as endomorphism algebras in certain monoidal subcategories JJ_\ell of the Brauer category. The structure and representation theory of KY algebras intertwine combinatorial, diagrammatic, and homological techniques, connecting directly to Kazhdan–Lusztig theory and alcove geometry, and generalizing classical semisimplicity paradigms through a family of generalized Chebyshev recurrences. Their study unifies a spectrum of algebraic and combinatorial objects, controlling both semisimple and non-semisimple phenomena.

1. Definition and Diagrammatic Presentation

Let n,0n, \ell \geq 0 (or 1\ell \geq -1) and KK a unital commutative ring (often a field). The Brauer algebra Brn(δ)\operatorname{Br}_n(\delta), parameterized by δK\delta \in K, has as basis the set of (n,n)(n,n) Brauer diagrams: pairings of $2n$ points, with multiplication realized diagrammatically by vertical concatenation and with a factor of δ\delta for each closed loop formed. The Kadar–Yu algebra J,n(δ)J_{\ell, n}(\delta) is defined as the span of those Brauer diagrams whose "left-height" does not exceed \ell, where left-height is the maximum, over all crossing points, of the minimal number of lines crossed from the face to the left boundary. This subspace is closed under multiplication, hence defines a subalgebra.

Algebraically, J,n(δ)J_{\ell,n}(\delta) is generated by:

  • Temperley–Lieb generators uiu_i (1i<n)(1 \leq i < n) satisfying the standard Temperley–Lieb relations:
    • ui2=δuiu_i^2 = \delta u_i,
    • uiui±1ui=uiu_i u_{i\pm 1} u_i = u_i,
    • uiuj=ujuiu_i u_j = u_j u_i for ij>1|i-j|>1.
  • Symmetric group generators sms_m (1m+1)(1 \leq m \leq \ell+1) satisfying Coxeter relations:
    • sm2=1s_m^2 = 1,
    • smsm+1sm=sm+1smsm+1s_m s_{m+1} s_m = s_{m+1} s_m s_{m+1},
    • smsk=sksms_m s_k = s_k s_m for mk>1|m-k|>1.
  • Mixed relations reflecting that sms_m acts on the first min(+2,n)\min(\ell+2, n) propagating lines:
    • smui=uisms_m u_i = u_i s_m unless mi1|m-i| \leq 1,
    • smumsm=um+1s_m u_m s_m = u_{m+1} (m=1,,)(m=1,\dots,\ell),
    • s+1ui=uis+1s_{\ell+1} u_i = u_i s_{\ell+1} for i>+1i > \ell+1.

Boundary cases: J1,n(δ)TLn(δ)J_{-1,n}(\delta) \cong \operatorname{TL}_n(\delta) is the Temperley–Lieb algebra, and Jn2,n(δ)=Brn(δ)J_{n-2,n}(\delta) = \operatorname{Br}_n(\delta) the full Brauer algebra. These algebras thus interpolate between Temperley–Lieb and Brauer as \ell varies.

2. Iterated Inflation Structure and Tower of Recollement

KY algebras naturally admit a tower structure as {J,n(δ)}n0\{J_{\ell,n}(\delta)\}_{n\geq 0}. Each algebra admits a filtration

0=I2I0I2In or n20 = I_{-2} \subset I_0 \subset I_2 \subset \dots \subset I_{n \text{ or } n-2}

where ImI_m is the span of diagrams with at most mm propagating lines. The successive quotients satisfy

Im/Im2VmKKSmKVmI_m / I_{m-2} \cong V_m \otimes_K K S_{m_\ell} \otimes_K V_m^*

with m=min(+2,m)m_\ell = \min(\ell+2, m) and VmV_m the space of upper half-diagrams of appropriate left-height with mm propagating lines. This realizes J,n(δ)J_{\ell,n}(\delta) as an iterated inflation algebra, successively inflating symmetric group algebras KSmK S_{m_\ell} by these diagram modules.

This tower structure satisfies the six axioms (A1–A6) of Cox–Martin–Parker–Xi (CMPX), defining a "tower of recollement" (ToR):

  • (A1) Localisation by idempotents reduces J,nJ_{\ell,n} to J,n2J_{\ell,n-2}.
  • (A2) Quasi-heredity with explicit heredity chains.
  • (A3) Embedding into algebras of higher nn by adding strands.
  • (A4) Bimodule isomorphisms relating J,nJ_{\ell,n} via idempotents to bimodules over J,n1,J,n2J_{\ell,n-1}, J_{\ell,n-2}.
  • (A5) Restriction of cell modules admits a filtration supported on (p±1,?)(p\pm1,?).
  • (A6) Each simple appears as a subfactor of a restriction from higher rank.

This realizes KY algebras as a ToR in the sense of [Cox–Martin–Parker–Xi, 2006] and controls the homological and cellular structure of representations (Alraddadi et al., 2024).

3. Standard Modules, Gram Determinants, and the Chebyshev Paradigm

Standard (cell) modules Δ(p,λ)n\Delta^{n}_{(p,\lambda)} of J,nJ_{\ell,n} are indexed by pairs (p,λ)(p,\lambda) where 0pn0 \leq p \leq n, npn-p even, and λmin(p,+2)\lambda \vdash \min(p, \ell+2). The construction involves a primitive idempotent in J,pJ_{\ell,p} and the diagrammatic module J(n,p)J_\ell(n,p), modulo diagrams with <p<p propagating lines. For p=np = n, the head standard module Δ(n,λ)n\Delta^{n}_{(n,\lambda)} coincides with the Specht module SλS_\lambda.

Cell modules admit a *-contravariant bilinear form (where * reverses diagrams). The Gram determinant Δ(p,λ)n\|\Delta^{n}_{(p,\lambda)}\| can be computed in terms of δ\delta or α\alpha, and for the important family p=n2p=n-2, these determinants satisfy a Chebyshev recurrence:

Pn+1(α)=αPn(α)+Pn1(α)P_{n+1}(\alpha) = \alpha P_n(\alpha) + P_{n-1}(\alpha)

Each partition λ+2\lambda \vdash \ell+2 yields a monic polynomial C(λ)(α)C^{(\lambda)}(\alpha) and reduced Chebyshev-type series Pn(λ)(α)P_n^{(\lambda)}(\alpha), such that for n+4n \geq \ell+4,

Δ(n2,λ)n=C(λ)(α)[Pn(λ)(α)]dimSλ.\|\Delta^{n}_{(n-2, \lambda)}\| = C^{(\lambda)}(\alpha)[P_n^{(\lambda)}(\alpha)]^{\dim S_\lambda}.

For example, for =1\ell=-1 (Temperley–Lieb), Pn()=Un1(α/2)P_n^{()} = U_{n-1}(\alpha/2) recovers the Chebyshev UU-polynomials, while for large \ell (Brauer), the determinants correspond to hook-length products (Morris et al., 31 Dec 2025).

Low-rank cases demonstrate the roles of partitions and the combinatorics of standard modules:

  • =0\ell=0: λ{(2),(12)}\lambda \in \{(2), (1^2)\}, each with explicit C(λ)C^{(\lambda)} and Chebyshev-series Pn(λ)P_n^{(\lambda)}.
  • =1\ell=1: λ\lambda runs over all partitions of $3$; similar explicit determinants arise.

4. Semisimplicity and Non-Semisimple Representation Theory

The semisimplicity of J,n(δ)J_{\ell,n}(\delta) over C\mathbb{C} is controlled by the Gram determinants. If δR\delta \notin \mathbb{R}, all Δ(p,λ)n0\|\Delta^{n}_{(p,\lambda)}\| \neq 0 for all p,λp,\lambda, so the algebra is semisimple. In this regime, every cell module is simple, labeled by (p,λ)(p,\lambda) as above, and the regular module decomposes accordingly:

J,n(δ)p,λEndK(Δ(p,λ))J_{\ell,n}(\delta) \cong \bigoplus_{p,\lambda} \operatorname{End}_K(\Delta(p,\lambda))

(Alraddadi et al., 2024).

For real δ\delta or specializations of α\alpha (e.g., roots of the associated Chebyshev series), the algebras may fail to be semisimple. At each vanishing determinant (a root α0\alpha_0 of some Pn(λ)P_n^{(\lambda)}), there exists a non-split map between standard modules, controlled by explicit "bootstrap" elements ξ(n)\xi^{(n)}. The structure of non-semisimple blocks, submodules, composition series, and socles is deterministic, exhibiting direct links to alcove geometry and representation-theoretic walls.

The general pattern in the non-semisimple case is controlled by the interplay between the tower functors (localisation and globalisation in the ToR) and the stratification of the module category via the Chebyshev determinants (Morris et al., 31 Dec 2025).

5. Alcove Geometry, Combinatorics, and the Bratteli/Rollet Graphs

KY algebras' module categories demonstrate a pronounced alcove-geometric structure. The indexing set (p,λ)(p,\lambda) organizes into the Rollet (projected Bratteli) graph, projecting the ordinary Bratteli diagram by omitting certain diagrammatic features ("ee-strands"). The head comprises partitions of size at most +2\ell+2, from which arms emanate corresponding to larger pp.

This labeling system can be embedded into Rr\mathbb{R}^r as the $1$-skeleton of a finite or affine Weyl alcove tessellation. For =1\ell=-1 (Temperley–Lieb), the set is one-dimensional and semisimplicity corresponds to the non-vanishing of a parameter outside the quantum-group root-of-unity locus 2cos(kπ/(n+1))2 \cos(k\pi/(n+1)). For higher \ell, the complexity escalates to higher-rank affine Weyl group reflection hyperplanes.

Non-semisimplicity loci (roots of Chebyshev series) correspond to reflection hyperplanes in α\alpha-space; the combinatorics of add/remove box moves on partitions (Young lattice) realize the percolation of submodules and the precise positioning of module filtrations. Exact sequences and composition factor structures are thus articulated geometrically (Morris et al., 31 Dec 2025).

6. Connections, Physical Motivation, and Low-Rank Examples

Kadar–Yu algebras are physically motivated as deformations and extensions of symmetric group and Temperley–Lieb representation theory in the setting of statistical mechanics, integrable models, and categorical constructions. They interpolate between Catalan (Temperley–Lieb) and double-factorial (Brauer) combinatorics, further connecting to Kazhdan–Lusztig theory via the interplay between cellular, quasi-hereditary, and topos-theoretic frameworks.

Low-rank instances explicitly demonstrate how the intermediate algebras possess richer representation theory than either extreme:

  • For n=4n=4, =0\ell=0, simples are
    • Δ(4,2)\Delta(4, \Box \vdash 2) (dim $2$),
    • Δ(2,2)\Delta(2, \Box \vdash 2) (dim $4$),
    • Δ(0,)\Delta(0, \emptyset) (dim $3$),
    • giving Br4(δ)M2(C)M4(C)M3(C)\operatorname{Br}_4(\delta) \cong M_2(\mathbb{C}) \oplus M_4(\mathbb{C}) \oplus M_3(\mathbb{C}) for non-real δ\delta.

The passage from =1\ell=-1 to =n2\ell=n-2 recovers the interpolation between Temperley–Lieb and Brauer from both diagrammatic and representation-theoretic standpoints, always controlled by the relevant determinant formulas and alcove geometry.

7. Summary Table: Special Cases

\ell Algebra Description Semisimplicity Condition
1-1 Temperley–Lieb No sjs_j generators δ{2cos(πk)}\delta \notin \{2\cos(\frac{\pi}{k})\}
$0$ "Rook–Brauer"/boundary-TL First additional crossings Non-real δ\delta; Pn(λ)(δ)0P_n^{(\lambda)}(\delta) \neq 0
n2n-2 Brauer Full diagram algebra Non-real δ\delta

Failures of semisimplicity correspond to roots of specialized Chebyshev polynomials Pn(λ)(α)=0P_n^{(\lambda)}(\alpha) = 0 and yield non-trivial homological and combinatorial phenomena.


Kadar–Yu algebras thus constitute a unifying and tractable family of diagrammatic algebras, whose module-theoretic and combinatorial properties capture a broad interpolation between well-understood classical examples, regulated by generalizations of the Chebyshev recurrence and alcove geometric framework (Alraddadi et al., 2024, Morris et al., 31 Dec 2025).

Definition Search Book Streamline Icon: https://streamlinehq.com
References (2)

Topic to Video (Beta)

No one has generated a video about this topic yet.

Whiteboard

No one has generated a whiteboard explanation for this topic yet.

Follow Topic

Get notified by email when new papers are published related to Kadar--Yu Algebras.