Syzygies of polymatroidal ideals

Yairon Cid-Ruiz, Jacob P. Matherne, Anna Shapiro · 2025

Read Syzygies of polymatroidal ideals by Yairon Cid-Ruiz, Jacob P. Matherne, Anna Shapiro (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-2507-13153.txt.

Introduction

Polymatroids, a broad family of rank-like structures, carry ideals whose algebraic shadows encode deep combinatorics. Yairon Cid-Ruiz and co-authors introduce a new polynomial for them in this 2025 paper, called the cave polynomial. It behaves well under combining structures, and its stretched form again traces a polymatroid.

The polynomial also gives a K-theory picture of the structure inside a larger ring built from a multisymmetric lift. Two open conjectures fall as applications: one about polymatroidal ideals themselves, and one about the Möbius support of a polymatroid. The title's syzygies, the relations among relations of the ideals, sit at the paper's heart. A single new invariant, the authors show, can organize several hard questions at once.

Catalog record

CREATORYairon Cid-Ruiz, Jacob P. Matherne, Anna Shapiro
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE37 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2507-13153

Browse the full library · About the Tower of Babel

Full text

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

Syzygies of polymatroidal ideals Yairon Cid-Ruiz, Jacob P. Matherne, Anna Shapiro Abstract . We introduce the cave polynomial of a polymatroid and show that it yields a valuative function on polymatroids. The support of this polynomial after homogenization is again a polymatroid. The cave polynomial gives a K K -theoretic description of a polymatroid in the augmented K K -ring of a multisymmetric lift. As applications, we settle two conjectures: one by Bandari, Bayati, and Herzog regarding polymatroidal ideals, and another by Castillo, Cid-Ruiz, Mohammadi, and Montaño regarding the Möbius support of a polymatroid. 1. Introduction A polymatroid \mathscr{P} on the set [p]=\{1,\ldots,p\} with cage {\mathbf{m}}=(m_{1},\ldots,m_{p})\in\mathbb{N}^{p} is given by a function \text{\rm rk}_{\mathscr{P}}\colon 2^{[p]}\rightarrow\mathbb{N} satisfying the following properties: (i) (Normalization) \text{\rm rk}_{\mathscr{P}}\left(\varnothing\right)=0 . (ii) (Monotonicity) \text{\rm rk}_{\mathscr{P}}\left(J_{1}\right)\leq\text{\rm rk}_{\mathscr{P}}\left(J_{2}\right) if J_{1}\subseteq J_{2}\subseteq[p] . (iii) (Submodularity) \text{\rm rk}_{\mathscr{P}}\left(J_{1}\cap J_{2}\right)+\text{\rm rk}_{\mathscr{P}}\left(J_{1}\cup J_{2}\right)\leq\text{\rm rk}_{\mathscr{P}}\left(J_{1}\right)+\text{\rm rk}_{\mathscr{P}}\left(J_{2}\right) for all J_{1},J_{2}\subseteq[p] . (iv) (Cage) \text{\rm rk}_{\mathscr{P}}\left(\{i\}\right)\leq m_{i} for all i\in[p] . We say that \text{\rm rk}_{\mathscr{P}}\colon 2^{[p]}\rightarrow\mathbb{N} is the rank function of \mathscr{P} and that the rank of \mathscr{P} is given by \text{\rm rk}(\mathscr{P})=\text{\rm rk}_{\mathscr{P}}([p]) . A polymatroid with cage {\mathbf{m}}=(1,\ldots,1) is called a matroid . Let R=\mathbb{k}[x_{1},\ldots,x_{p}] be a polynomial ring over a field \mathbb{k} . Let \mathscr{P} be a polymatroid on the set [p] with cage {\mathbf{m}}\in\mathbb{N}^{p} . The polymatroidal ideal I_{\mathscr{P}}\subset R of \mathscr{P} is the monomial ideal generated by the monomials corresponding to the lattice points in the base polytope B(\mathscr{P}) of \mathscr{P} . For each i\geq 0 , the i -th homological shift ideal {\rm HS}_{i}(I_{\mathscr{P}})\subset R of I_{\mathscr{P}} is the monomial ideal generated by the monomials corresponding to the shifts in the i -th position of the minimal free R -resolution of I_{\mathscr{P}} . Let I(\mathscr{P}) be the independence polytope of \mathscr{P} . The Möbius function \mu_{\mathscr{P}}\colon\mathbb{Z}^{p}\rightarrow\mathbb{Z} of the polymatroid \mathscr{P} is defined inductively by setting \mu_{\mathscr{P}}({\mathbf{n}})=1 if {\mathbf{n}}\in B(\mathscr{P}) and \mu_{\mathscr{P}}({\mathbf{n}})\;=\;1-\sum_{\mathbf{w}\,\in\,\left({\mathbf{n}}+\mathbb{Z}_{>0}^{p}\right)\cap I(\mathscr{P})}\mu_{\mathscr{P}}(\mathbf{w}) if {\mathbf{n}}\in I(\mathscr{P})\setminus B(\mathscr{P}) . For all {\mathbf{n}}\in\mathbb{Z}^{p}\setminus I(\mathscr{P}) , we set \mu_{\mathscr{P}}({\mathbf{n}})=0 . The Möbius support of \mathscr{P} is defined as \mu{\text{\rm-supp}}(\mathscr{P})=\{{\mathbf{n}}\in\mathbb{N}^{p}\mid\mu_{\mathscr{P}}({\mathbf{n}})\neq 0\} . The main goal of this paper is to settle the following two conjectures regarding polymatroids. Conjecture 1.1 (Bandari – Bayati – Herzog [ Bay18 , HMRZ21 ] ) . All the homological shift ideals {\rm HS}_{i}(I_{\mathscr{P}}) of I_{\mathscr{P}} are again polymatroidal ideals. Conjecture 1.2 (Castillo – Cid-Ruiz – Mohammadi – Montaño [ CCRMM22 ] ) . The Möbius support of \mathscr{P} is a generalized polymatroid ( i.e., a homogenization of it yields a polymatroid ) . Section 1 has been verified in the following cases: in [ Bay18 ] , if \mathscr{P} is a matroid; in [ HMRZ21 ] , if \mathscr{P} satisfies the strong exchange property; in [ FH23 ] , if \mathscr{P} has rank two; see also [ Fic22 ] . In [ CCRMM22 , Theorem 7.17] , the conclusion of Section 1 was proven in the case where \mathscr{P} is realizable, thus serving as motivation to state this conjecture. By [ CCRMM22 , Theorem 7.19] or [ EL23 , Remark 3.5] , we know that Section 1 holds when \mathscr{P} is a matroid. The K -ring of a matroid was recently introduced by Larson, Li, Payne, and Proudfoot [ LLPP24 ] . Since the K -ring of a matroid has already become an object of interest, we are also interested in a K -theoretic description of the polymatroid \mathscr{P} . Let \mathscr{M} be a matroid on a ground set E with subsets \mathcal{S}_{1},\ldots,\mathcal{S}_{p}\subseteq E such that the restriction polymatroid is \mathscr{P} . By considering the augmented K -ring of \mathscr{M} , we say that the Snapper polynomial of \mathscr{P} is given by \text{\rm Snapp}_{\mathscr{P}}\left(t_{1},\ldots,t_{p}\right)\;=\;\chi\left(\mathscr{M},\,\mathcal{L}_{\mathcal{S}_{1}}^{\otimes t_{1}}\otimes\cdots\otimes\mathcal{L}_{\mathcal{S}_{p}}^{\otimes t_{p}}\right). For more details, see Section 2 , Section 2 , and Section 2 . Motivated by the combinatorial notion of caves introduced in [ CCRMM22 ] , we introduce the cave polynomial of a polymatroid. The cave polynomial of \mathscr{P} is given by \text{\rm cave}_{\mathscr{P}}(t_{1},\ldots,t_{p})\;=\;\sum_{{\mathbf{n}}\in\mathbb{N}^{p}\text{ and }|{\mathbf{n}}|=\text{\rm rk}(\mathscr{P})}\mathbb{1}_{\mathscr{P}}({\mathbf{n}})\,\prod_{i=1}^{p-1}\left(1-\max_{i<j}{\big\{\mathbb{1}_{\mathscr{P}}\left({\mathbf{n}}-{\mathbf{e}}_{i}+{\mathbf{e}}_{j}\right)\big\}}t_{i}^{-1}\right){\mathbf{t}}^{\mathbf{n}}, where \mathbb{1}_{\mathscr{P}} denotes the indicator function of the base polytope B(\mathscr{P}) of \mathscr{P} . It turns out that the Snapper polynomial \text{\rm Snapp}_{\mathscr{P}}(t_{1},\ldots,t_{p}) and the cave polynomial \text{\rm cave}_{\mathscr{P}}(t_{1},\ldots,t_{p}) encode the same information. Indeed, we have the equality \text{\rm Snapp}_{\mathscr{P}}(t_{1},\ldots,t_{p})\;=\;\mathfrak{b}\big(\text{\rm cave}_{\mathscr{P}}(t_{1},\ldots,t_{p})\big), where \mathfrak{b}\colon\mathbb{Q}[t_{1},\ldots,t_{p}]\rightarrow\mathbb{Q}[t_{1},\ldots,t_{p}] is the \mathbb{Q} -linear map sending t_{1}^{n_{1}}\cdots t_{p}^{n_{p}} to \binom{t_{1}+n_{1}}{n_{1}}\cdots\binom{t_{p}+n_{p}}{n_{p}} (see (3) ). Our goal is to investigate various aspects of the cave polynomial. When \mathscr{P} is realizable, our approach is to consider the corresponding multiplicity-free variety (see Section 2 ). To address the general case (where \mathscr{P} need not be realizable), our main idea is to show that the cave polynomial yields a valuative function on polymatroids. The theorem below contains our main results. Theorem A . Section 1 and Section 1 hold. More precisely, we have: (i) The support of the cave polynomial \text{\rm cave}_{\mathscr{P}}(t_{1},\ldots,t_{p}) of \mathscr{P} is a generalized polymatroid. (ii) The cave polynomial \text{\rm cave}_{\mathscr{P}}(t_{1},\ldots,t_{p}) of \mathscr{P} satisfies the equality \text{\rm cave}_{\mathscr{P}}(t_{1},\ldots,t_{p})\;=\;\sum_{{\mathbf{n}}\in\mathbb{N}^{p}}\,\mu_{\mathscr{P}}({\mathbf{n}})\,t_{1}^{n_{1}}\cdots t_{p}^{n_{p}}. In particular, Section 1 holds. (iii) The K -polynomial of the polymatroidal ideal I_{\mathscr{P}}\subset R is given by \mathcal{K}\left(I_{\mathscr{P}};t_{1},\ldots,t_{p}\right)\;=\;t_{1}^{m_{1}}\cdots t_{p}^{m_{p}}\;\text{\rm cave}_{\mathscr{P}^{\vee}}\left(t_{1}^{-1},\ldots,t_{p}^{-1}\right), where \mathscr{P}^{\vee}={\mathbf{m}}-\mathscr{P} is the dual polymatroid with respect to the cage {\mathbf{m}} . Thus the i -th homological shift ideal of I_{\mathscr{P}} is given by {\rm HS}_{i}\left(I_{\mathscr{P}}\right)\;=\;\Big(x_{1}^{n_{1}}\cdots x_{p}^{n_{p}}\;\mid\;{\mathbf{n}}\in\mathbb{N}^{p},\;|{\mathbf{n}}|=\text{\rm rk}(\mathscr{P})+i\text{\, and \,}\mu_{\mathscr{P}^{\vee}}({\mathbf{m}}-{\mathbf{n}})\neq 0\Big). In particular, Section 1 holds. (iv) The function \mathscr{P}\mapsto\text{\rm cave}_{\mathscr{P}}(t_{1},\ldots,t_{p}) assigning the cave polynomial to a polymatroid is valuative. 2. Proofs of our results Let \mathscr{P} be a polymatroid on [p]=\{1,\ldots,p\} with rank function \text{\rm rk}_{\mathscr{P}}\colon 2^{[p]}\rightarrow\mathbb{Z} . Let {\mathbf{m}}=(m_{1},\ldots,m_{p})\in\mathbb{N}^{p} be a cage for the polymatroid \mathscr{P} . This means that \text{\rm rk}_{\mathscr{P}}(\{i\})\;\leq\;m_{i}\quad\text{ for all }\quad 1\leq i\leq p. Let \mathbb{k} be a field and R=\mathbb{k}\left[x_{1},\ldots,x_{p}\right] be a standard \mathbb{N}^{p} -graded polynomial ring with \deg(x_{i})={\mathbf{e}}_{i}\in\mathbb{N}^{p} for every i . Let S=\mathbb{k}\left[x_{i,j}\mid 1\leq i\leq p,\,0\leq j\leq m_{i}\right] be a standard \mathbb{N}^{p} -graded polynomial ring with \deg(x_{i,j})={\mathbf{e}}_{i}\in\mathbb{N}^{p} for every i,j . We note that \text{MultiProj}(S)\;=\;{\mathbb{P}}\;:=\;{\mathbb{P}}_{\mathbb{k}}^{m_{1}}\times_{\mathbb{k}}\cdots\times_{\mathbb{k}}{\mathbb{P}}_{\mathbb{k}}^{m_{p}} is the product of projective spaces associated to S . The base polytope of the polymatroid \mathscr{P} is given by B(\mathscr{P})\;:=\;\Big\{\begin{array}[]{c|c}\mathbf{v}=(v_{1},\ldots,v_{p})\in\mathbb{R}_{\geq 0}^{p}&\;\;\sum_{i=1}^{p}v_{i}=\text{\rm rk}(\mathscr{P})\;\text{ and }\;\sum_{j\in J}v_{j}\leq\text{\rm rk}(J)\text{ for all }J\subseteq[p]\end{array}\Big\}. The independence polytope of \mathscr{P} is defined as I(\mathscr{P})\;:=\;\Big\{\begin{array}[]{c|c}\mathbf{v}=(v_{1},\ldots,v_{p})\in\mathbb{R}_{\geq 0}^{p}&\;\;\sum_{j\in J}v_{j}\leq\text{\rm rk}(J)\text{ for all }J\subseteq[p]\end{array}\Big\}. We have the following equality I(\mathscr{P})\;=\;\Big(B(\mathscr{P})\;+\;\mathbb{R}_{\leq 0}^{p}\Big)\;\cap\;\mathbb{R}_{\geq 0}^{p}, where + denotes the Minkowski sum. Our two objects of interest are the following. Definition 2.1 . (i) The polymatroidal ideal I_{\mathscr{P}}\subset R of the polymatroid \mathscr{P} is the monomial ideal given by I_{\mathscr{P}}\;:=\;\left({\mathbf{x}}^{\mathbf{n}}=x_{1}^{n_{1}}\cdots x_{p}^{n_{p}}\,\mid\,{\mathbf{n}}\in B(\mathscr{P})\cap\mathbb{N}^{p}\right). (ii) The Möbius function \mu_{\mathscr{P}}\colon\mathbb{Z}^{p}\rightarrow\mathbb{Z} of the polymatroid \mathscr{P} is defined inductively by setting \mu_{\mathscr{P}}({\mathbf{n}}):=1 if {\mathbf{n}}\in B(\mathscr{P}) and \mu_{\mathscr{P}}({\mathbf{n}})\;:=\;1-\sum_{\mathbf{w}\,\in\,\left({\mathbf{n}}+\mathbb{Z}_{>0}^{p}\right)\cap I(\mathscr{P})}\mu_{\mathscr{P}}(\mathbf{w}) if {\mathbf{n}}\in I(\mathscr{P})\setminus B(\mathscr{P}) . When {\mathbf{n}}\not\in I(\mathscr{P}) , we set \mu_{\mathscr{P}}({\mathbf{n}}):=0 . Then the Möbius support of \mathscr{P} is defined as \mu{\text{\rm-supp}}(\mathscr{P})\;:=\;\big\{{\mathbf{n}}\in\mathbb{N}^{p}\,\mid\,\mu_{\mathscr{P}}({\mathbf{n}})\neq 0\big\}. Consider the minimal \mathbb{Z}^{p} -graded free R -resolution \mathbb{F}_{\bullet}\colon\quad\cdots\rightarrow\;F_{i}=\bigoplus_{j=1}^{\beta_{i}}R(-\mathbf{b}_{i,j})\;\rightarrow\cdots\rightarrow\;F_{0}\;\rightarrow\;I_{\mathscr{P}}\;\rightarrow 0 of I_{\mathscr{P}} , where each \mathbf{b}_{i,j}=(b_{i,j,1},\ldots,b_{i,j,p})\in\mathbb{N}^{p} . The i -th homological shift ideal of I_{\mathscr{P}} is given by {\rm HS}_{i}(I_{\mathscr{P}})\;:=\;\left({\mathbf{x}}^{\mathbf{b}_{i,j}}\mid 1\leq j\leq\beta_{i}\right)\;\subset\;R. Notice that the equality {\rm HS}_{0}(I_{\mathscr{P}})=I_{\mathscr{P}} holds. Definition 2.2 . The K -polynomial of I_{\mathscr{P}} is defined as \mathcal{K}(I_{\mathscr{P}};t_{1},\ldots,t_{p})\;:=\;\sum_{i\geq 0}{(-1)}^{i}\sum_{j=1}^{\beta_{i}}{\mathbf{t}}^{\mathbf{b}_{i,j}}\;\in\;\mathbb{Z}[t_{1},\ldots,t_{p}] (see [ MS05 ] , [ KM05 ] ). Remark 2.3 . By an abuse of notation, we also denote by \mathscr{P} the associated base discrete polymatroid (i.e., the lattice points in B(\mathscr{P})\cap\mathbb{N}^{p} ). Being a base discrete polymatroid is equivalent to being an

Read the complete text