Cohomological support varieties for monomial ideals
Kara Fagerstrom, Julianne Faur, Benjamin Katz, Kesavan Mohana Sundaram, Stephen Stern, Ryan Watson · 2026
Read Cohomological support varieties for monomial ideals by Kara Fagerstrom, Julianne Faur, Benjamin Katz, Kesavan Mohana Sundaram, Stephen Stern, Ryan Watson (2026) 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-2605-29103.txt.
Introduction
Kara Fagerstrom classifies cohomological support varieties for monomial ideals with six generators. Earlier work had covered up to five generators; this paper extends the work to six. The authors also build two families of rings. One yields unbounded codimension, and the other yields an unbounded count of components. Finally, they answer a question of Gintz about which ones occur when the ideal is the edge ideal of a cycle. The results map the full range of shapes these objects can take in this case.
Catalog record
| CREATOR | Kara Fagerstrom, Julianne Faur, Benjamin Katz, Kesavan Mohana Sundaram, Stephen Stern, Ryan Watson |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 133880 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.29103
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Cohomological support varieties for monomial ideals Kara Fagerstrom, Julianne Faur, Benjamin Katz, Kesavan Mohana Sundaram, Stephen Stern, Ryan Watson Abstract . Let R R be a local or positively graded ring with a regular presentation R ≅ Q / I R\cong Q/I where I I is a monomial ideal generated by n n elements on a regular sequence. In [ 9 ] the authors classify the cohomological support varieties 𝒱 R ( R ) {\mathcal{V}}_{R}(R) for n ⩽ 5 n\leqslant 5 . In this paper we extend their results to classify the varieties that can occur as 𝒱 R ( R ) {\mathcal{V}}_{R}(R) for n = 6 n=6 . Moreover, we provide two families of rings, one realizing cohomological support varieties of unbounded codimension, the other realizing an unbounded number of components. Finally, we answer a question of [ 12 ] about the varieties that occur as 𝒱 R ( R ) {\mathcal{V}}_{R}(R) where I I is given by the edge ideal of a cycle. 2020 Mathematics Subject Classification Primary: 13D99. Secondary: 05E40 Introduction Inspired by Quillen’s geometric approach to modular representation theory [ 19 ] , Avramov imported the theory of support varieties to commutative algebra to study local complete intersection rings [ 5 ] . This has now expanded to include all noetherian local rings through the work of various authors [ 14 , 10 , 17 , 18 ] . Most notably for this paper, Pollitz developed the theory of cohomological support varieties over Koszul complexes which generalizes the theory of Avramov. Suppose that R is a local or positively graded ring. Then R is said to admit a minimal regular presentation, if one of the following holds: • \widehat{R}=Q/I with (Q,\mathfrak{m},k) a regular local ring and I\subset\mathfrak{m}^{2} in the local case, or • R=Q/I where Q is a positively graded polynomial algebra over a field k and I is an ideal generated by homogeneous forms of degree at least 2 in the graded case. Assume that R has a minimal regular presentation given by Q/I , where I is minimally generated by n elements. Let E be the Koszul complex on a minimal generating set of I . The ring of cohomological operators \mathcal{S}=k[\chi_{1},\dotsc,\chi_{n}] is a polynomial ring over k where each \chi_{i} is a variable of cohomological degree 2 . For any finitely generated R -module M , the ring \mathcal{S} acts on {\operatorname{Ext}}_{E}(M,k) , making {\operatorname{Ext}}_{E}(M,k) into a graded \mathcal{S} -module. There are many different ways to define this action, but one is explained as follows. The ring \mathcal{S} is a graded k -subalgebra of {\operatorname{Ext}}_{E}(k,k) , and the left action of {\operatorname{Ext}}_{E}(k,k) on {\operatorname{Ext}}_{E}(M,k) defines an action of \mathcal{S} on {\operatorname{Ext}}_{E}(M,k) (see [ 1 , Section 2] ). The specific details of how \mathcal{S} acts on this {\operatorname{Ext}}_{E}(M,N) are not needed for this paper, so we refer the interested reader to [ 17 , 3.2.6] . Due to our conditions on Q , it turns out that {\operatorname{Ext}}_{E}(M,k) is a finitely generated graded \mathcal{S} module [ 17 , 3.2.5] . The cohomological support variety of a finitely generated R module M , is defined to be {\mathcal{V}}_{R}(M):={\operatorname{Supp}}_{\mathcal{S}}\left({\operatorname{Ext}}_{E}(M,k)\right). As {\operatorname{Ext}}_{E}(M,k) is a finitely generated graded \mathcal{S} module, we can and will view {\mathcal{V}}_{R}(M) as a conical affine variety (i.e. an affine cone), living in \mathbb{A}^{n}_{k} which we identify with \mathfrak{m}{\operatorname{Spec}}(\mathcal{S}) . Geometric properties of these varieties encode homological information of both the ring and the module. For example, they see some of the structure of the lattice of thick subcategories in the derived category of R [ 20 , 17 ] . When the ring is a complete intersection, the theory of support varieties is generally well-understood and has been studied extensively [ 5 , 2 , 6 , 4 , 11 ] . However, moving away from the complete intersection case, their behavior becomes much more mysterious. In [ 18 ] , the author asks the following question known as the realization problem for cohomological support varieties: Question 1 . Given a noetherian local ring R , what conical affine varieties in \mathbb{A}^{n}_{k} can be realized as the support variety of some finitely generated R -module? This problem is solved when the ring is a complete intersection [ 6 , 4 ] , with the answer being that every conical affine variety is realizable. However, when the ring is not a complete intersection, much less is known. There are bounds on the dimensions of the varieties that can occur [ 8 , 9 ] , and certain varieties can be guaranteed to be realizable over certain rings in [ 18 , 7 , 22 , 12 ] . Beyond these results, little is known in general. In fact, it is unknown what varieties can arise as {\mathcal{V}}_{R}(R) . Some results in this direction include the following. When R is a complete intersection, {\mathcal{V}}_{R}(R)=\{0\} [ 5 ] , and when R is a Golod ring of codepth at least two, {\mathcal{V}}_{R}(R) is all of \mathbb{A}^{n}_{k} [ 8 ] . Moreover, the possible support varieties arising from {\mathcal{V}}_{R}(R) when R has small codepth is contained in [ 18 ] . In [ 9 ] the authors classify all varieties that can be realized as {\mathcal{V}}_{R}(R) when R is defined by a monomial ideal generated by five elements on a regular sequence of Q . These are exactly linear subspaces with dimension not 1 and the union of two hyperplanes. In [ 12 ] , the author computes the support varieties of equigenerated monomial ideals with six generators on a regular sequence of Q when the residue field is \mathbb{Q} . In this paper, we extend the results of [ 9 ] and [ 12 ] to all rings defined by six monomial relations with no restrictions on the residue field. Theorem A . Let R be a local or positively graded ring with minimal regular presentation given by Q/I where I is a monomial ideal generated by six elements on a regular sequence of Q . Then up to a reordering of the monomial generators of I , the varieties realized as {\mathcal{V}}_{R}(R) are listed below: • a coordinate subspace of \mathbb{A}^{6}_{k} with dimension not equal to 1 • V(\chi_{1},\;\chi_{4}\chi_{6}) • V(\chi_{4}\chi_{6}) • V(\chi_{4}\chi_{6},\;\chi_{5}\chi_{6}) • V(\chi_{2}\chi_{4}\chi_{6}) • V(\chi_{1}\chi_{3}\chi_{5}+\chi_{2}\chi_{4}\chi_{6}) . When dealing with rings defined by monomial ideals, one has to go all the way to five generators to get a ring whose support variety is not a linear subspace [ 9 ] . The only such variety is given by V(\chi_{1}\chi_{5}) which is a union of hyperplanes which can be realized by R=k\llbracket x,y,z,w\rrbracket/I where I=(x^{2},xy,yz,zw,w^{2}) . Our result shows that one need only go up to six generators to attain a non-linear hypersurface, namely V(\chi_{1}\chi_{3}\chi_{5}+\chi_{2}\chi_{4}\chi_{6}) . Expanding beyond this variety, we obtain the following theorem which answers a question of [ 12 ] , about the support variety of R where R is defined by the edge ideal of a cycle. Theorem B . Let R=k[x_{1},\dotsc,x_{n}]/(x_{1}x_{2},x_{2}x_{3},\dotsc,x_{n}x_{1}) , the ring defined by the edge ideal of an n -cycle. Then {\mathcal{V}}_{R}(R)=\begin{cases}V(\chi_{1}\chi_{3}\ldots\chi_{n-1}+\chi_{2}\chi_{4}\ldots\chi_{n})&\text{if}\ n\equiv 2\pmod{4}\\ \mathbb{A}_{k}^{n}&otherwise.\end{cases} Note that specializing to the case n=6 gives the non-linear hypersurface mentioned above. The structure of the paper