Gateva-Ivanova and Majid (2024) proved that the quadratic algebra A(k,X,r)A(k,X,r) of a finite left-nondegenerate idempotent solution of the set-theoretic Yang-Baxter equation is always a PBW algebra of Gelfand-Kirillov dimension 1 with dim⁡kAd=n=∣X∣\dim_k A_d = n = |X| for every dd, and studied its Veronese subalgebras through an infinite prolongation sequence of new solutions (X,r(d))(X, r^{(d)}) on the same set. The paper explicitly poses two open questions. Question 4.12: for a left-nondegenerate idempotent quadratic set (braid relations not assumed) whose quadratic algebra is PBW with the fixed enumeration and has Gelfand-Kirillov dimension 1, what is the exact upper bound for dim⁡kA2\dim_k A_2, equivalently the minimal possible number of quadratic relations? Only the bounds n≤dim⁡kA2≤(n2)+1n \le \dim_k A_2 \le \binom{n}{2}+1 are known. Question 6.12: can a braided set be characterised by properties of its prolongation sequence, and in the left-nondegenerate idempotent case what is the stopping distance (the least d−1d-1 with (X,r(d))=(X,r)(X,r^{(d)})=(X,r)) and how many distinct solutions arise?

Background

The Yang-Baxter equation for a linear operator R:V⊗V→V⊗VR:V\otimes V\to V\otimes V underlies quantum groups and braid representations; its set-theoretic analogue was proposed in the early 1990s and systematically studied from Etingof-Schedler-Soloviev onwards. A braided set (or set-theoretic solution of the Yang-Baxter equation, YBE) is a pair (X,r)(X,r) with a map r:X×X→X×Xr:X\times X\to X\times X satisfying the braid relations r12r23r12=r23r12r23r^{12}r^{23}r^{12}=r^{23}r^{12}r^{23} on X3X^3, where r12=r×idXr^{12}=r\times\mathrm{id}_X and r23=idX×rr^{23}=\mathrm{id}_X\times r. Writing r(x,y)=(xy,xy)r(x,y)=({}^{x}y, x^{y}), the set is left-nondegenerate when Lx:y↦xyL_x:y\mapsto{}^{x}y is bijective for each xx, involutive when r2=idr^2=\mathrm{id}, and idempotent when r2=rr^2=r. The involutive nondegenerate case is algebraically well understood, while the strictly braided (non-involutive) case is far less developed; idempotent solutions form a non-involutive class studied for its own sake.

Gateva-Ivanova and Majid work with the more general notion of a quadratic set: any pair (X,r)(X,r) with r:X×X→X×Xr:X\times X\to X\times X, no braid relations or bijectivity assumed. To a finite quadratic set and a field kk one associates a quadratic algebra (an algebra generated in degree 1 with only degree-2 relations) A(k,X,r)=k⟨X⟩/(ℜA)A(k,X,r)=k\langle X\rangle/(\Re_A): fixing an enumeration X={x1,…,xn}X=\{x_1,\dots,x_n\} and the degree-lexicographic order, the relations rewrite each word xy∈X2xy\in X^2 to a degree-lexicographically smaller word of its orbit under the action of rr on X2X^2, so rr encodes a canonically presented monoid algebra. When (X,r)(X,r) solves the YBE this is called its Yang-Baxter algebra, a set-theoretic analogue of the quantum plane. A quadratic algebra is PBW (Poincare-Birkhoff-Witt type) when its relations extend to a quadratic Grobner basis, so that the normal words (those avoiding the leading monomials of the Grobner basis) form a kk-basis; its Gelfand-Kirillov dimension (GKdim) measures the growth of the dimensions of its graded components, with GKdim 1 meaning linear growth.

The 2024 source paper proves that for every finite left-nondegenerate idempotent solution of the YBE the relations ℜA\Re_A are themselves the reduced Grobner basis, hence A(k,X,r)A(k,X,r) is PBW with normal words {x1d−1xj}\{x_1^{d-1}x_j\} and dim⁡kAd=n\dim_k A_d=n for all d≥1d\ge 1 (Hilbert series (1+(n−1)t)/(1−t)(1+(n-1)t)/(1-t), GKdim 1), and combines this with a new graph-theoretic theorem on general PBW algebras to deduce infinite global dimension for these algebras. The combinatorics is carried by the directed graph ΓN\Gamma_N whose arrows are the length-2 normal words, so dim⁡kA2\dim_k A_2 counts arrows; under GKdim 1 the non-loop part of this graph is a subgraph of an acyclic tournament (an orientation of the complete graph on the nn vertices containing no directed cycle) with at most one self-arrow, giving dim⁡kA2≤(n2)+1\dim_k A_2\le\binom{n}{2}+1, while left-nondegeneracy and idempotence force dim⁡kA2≥n\dim_k A_2\ge n. Crucially, these bounds concern quadratic sets that need not satisfy the braid relations: for actual YBE solutions the value is pinned at nn, and a quadratic monomial algebra on variables y1,…,yny_1,\dots,y_n whose length-2 normal words are {yiyj∣i<j}∪{y1y1}\{y_iy_j\mid i<j\}\cup\{y_1y_1\} attains (n2)+1\binom{n}{2}+1, but whether the upper bound or any intermediate value is attained by the algebra of a left-nondegenerate idempotent quadratic set is unknown.

The paper also develops Veronese theory: the dd-Veronese subalgebra A(d)A^{(d)} of a graded algebra is the subalgebra spanned by degrees divisible by dd, the classical projective-geometry construction whose Yang-Baxter-algebra analogue was built earlier for involutive solutions. In the left-nondegenerate idempotent case the authors show that the set NdN_d of normal words of degree dd carries a canonical left-nondegenerate idempotent solution (Nd,ρd)(N_d,\rho_d) with Nd≅XN_d\cong X, and that A(k,X,r)(d)≅A(k,X,r(d))A(k,X,r)^{(d)}\cong A(k,X,r^{(d)}) for a new solution (X,r(d))(X,r^{(d)}) on the same underlying set; iterating gives the prolongation sequence of (X,r)(X,r). For the permutation idempotent solutions rf(x,y)=(f(y),y)r_f(x,y)=(f(y),y) studied in the precursor work of the same authors, the sequence is periodic with period the order of ff (indeed A(d)≅A(k,X,rfd)A^{(d)}\cong A(k,X,r_{f^d})), and the paper gives a 3-point example whose prolongation sequence alternates between rr and r(2)r^{(2)}; no general description of the periodicity, the return to rr, or the number of distinct solutions in the sequence is known.

Problem Statement

The problem consists of the two open questions explicitly posed by Gateva-Ivanova and Majid in arXiv:2409.02939, which share one objective: replace, by exact results, the bounds and partial descriptions that paper establishes for left-nondegenerate idempotent sets and their quadratic algebras. Both deliverables are read off the same normal-word combinatorics of A(k,X,r)A(k,X,r).

Question 4.12 (exact relation count). Let (X,r)(X,r) be a left-nondegenerate idempotent quadratic set on X={x1,…,xn}X=\{x_1,\dots,x_n\} (that is, r2=rr^2=r and each Lx:y↦xyL_x:y\mapsto{}^{x}y is bijective; the braid relations are not assumed). Suppose the associated quadratic algebra A=A(k,X,r)A=A(k,X,r) is PBW with PBW-generators x1,…,xnx_1,\dots,x_n in this fixed enumeration, and that GKdim A=1\mathrm{GKdim}\,A=1. Determine the exact upper bound for dim⁡kA2\dim_k A_2, equivalently the minimal possible number of quadratic relations of AA, as a function of nn. The known bounds are n≤dim⁡kA2≤(n2)+1n\le\dim_k A_2\le\binom{n}{2}+1; the lower bound is attained (every left-nondegenerate idempotent solution of the YBE has dim⁡kA2=n\dim_k A_2=n), and it is open whether the upper bound, or any intermediate value, is attained within the stated class.

Question 6.12 (prolongation dynamics). (1) Can a braided set (X,r)(X,r) be characterised by properties of its prolongation sequence (X,r)(d)=(Nd,ρd)(X,r)^{(d)}=(N_d,\rho_d), that is, is there a necessary-and-sufficient condition on the prolongation sequence for (X,r)(X,r) to satisfy the braid relations? For a general braided set the underlying sets in general vary with dd; in the left-nondegenerate idempotent case the prolongations refer uniformly back to the same set XX, giving solutions (X,r(d))(X,r^{(d)}). (2) In the left-nondegenerate idempotent case, what is the stopping distance, defined as d−1d-1 for the smallest dd such that the dd-prolongation (X,r(d))(X,r^{(d)}) returns to (X,r)(X,r), and how many distinct solutions arise in the sequence? For permutation idempotent solutions (X,rf)(X,r_f) the sequence is periodic with period the order of ff, and one 3-point example alternates between rr and r(2)r^{(2)}; a general description of the periodicity and of the set of distinct solutions occurring is not known.

A complete answer identifies the exact maximal value of dim⁡kA2\dim_k A_2 (with proof of the universal upper bound over the stated class and explicit attaining examples) and establishes the periodicity structure of prolongation sequences for left-nondegenerate idempotent solutions (stopping distance and number of distinct solutions), settling the characterization part of Question 6.12 either by a theorem or by explicit counterexamples.

References

  1. Tatiana Gateva-Ivanova and Shahn Majid, Quadratic algebras and idempotent braided sets, preprint arXiv:2409.02939 (2024), DOI 10.48550/arXiv.2409.02939, https://arxiv.org/abs/2409.02939
  2. Tatiana Gateva-Ivanova and Shahn Majid, Quadratic algebras associated to permutation idempotent solutions of the YBE, preprint arXiv:2308.11427 (2023), DOI 10.48550/arXiv.2308.11427, https://arxiv.org/abs/2308.11427
  3. Tatiana Gateva-Ivanova, Veronese subalgebras and Veronese morphisms for a class of Yang-Baxter algebras, Journal of Noncommutative Geometry 20 (2025), 221-267, DOI 10.4171/jncg/626, https://doi.org/10.4171/jncg/626 (preprint arXiv:2204.08850, https://arxiv.org/abs/2204.08850)
  4. P. Etingof, T. Schedler and A. Soloviev, Set-theoretical solutions to the quantum Yang-Baxter equation, Duke Mathematical Journal 100 (1999), no. 2, 169-209, DOI 10.1215/s0012-7094-99-10007-x, https://doi.org/10.1215/s0012-7094-99-10007-x (preprint arXiv:math/9801047, https://arxiv.org/abs/math/9801047)
  5. Anastasia Doikou, Marzia Mazzotta and Paola Stefanelli, Parametric reflection maps: an algebraic approach, Communications in Algebra (published online 2026), DOI 10.1080/00927872.2026.2621983, https://doi.org/10.1080/00927872.2026.2621983 (preprint arXiv:2412.15839, https://arxiv.org/abs/2412.15839)
  6. Anastasia Doikou, Self-distributivity, braces, and the Yang-Baxter equation, in: 2024 MATRIX Annals, Part II, MATRIX Book Series, Springer (2026), DOI 10.1007/978-3-032-16206-9_4, https://doi.org/10.1007/978-3-032-16206-9_4 (preprint arXiv:2409.20479 under the title Self-distributive structures, braces & the Yang-Baxter equation, https://arxiv.org/abs/2409.20479)
  7. I. Colazzo, E. Jespers, L. Kubat and A. Van Antwerpen, Structure algebras of finite set-theoretic solutions of the Yang-Baxter equation, Journal of Algebra 705 (2026), 61-117, DOI 10.1016/j.jalgebra.2026.04.044, https://doi.org/10.1016/j.jalgebra.2026.04.044 (preprint arXiv:2305.06023, https://arxiv.org/abs/2305.06023)
  8. I. Colazzo, E. Jespers, L. Kubat, A. Van Antwerpen and C. Verwimp, Finite idempotent set-theoretic solutions of the Yang-Baxter equation, International Mathematics Research Notices 2024 (2024), no. 7, 5458-5489, DOI 10.1093/imrn/rnad183, https://doi.org/10.1093/imrn/rnad183 (preprint arXiv:2212.07361, https://arxiv.org/abs/2212.07361)
  9. Pavlos Kassotakis and Maciej Nieszporski, Idempotent compatible maps and discrete integrable systems on the triangular lattice, preprint arXiv:2504.12212 (2025), DOI 10.48550/arXiv.2504.12212, https://arxiv.org/abs/2504.12212