---
title: Saturated Newton Polytope Property
url: https://www.emergentmind.com/topics/saturated-newton-polytope-property
type: topic
---

# Saturated Newton Polytope Property

The **saturated Newton polytope property** (SNP) is a support-saturation condition for a polynomial or Laurent polynomial: every lattice point of the convex hull of its exponent vectors must itself occur as the exponent of a nonzero monomial. In standard notation, if \(f=\sum_\alpha c_\alpha x^\alpha\), then SNP means \(\Newton(f)\cap\mathbb Z^n=\Supp(f)\), where \(\Supp(f)=\{\alpha:c_\alpha\neq0\}\) and \(\Newton(f)=\Conv(\Supp(f))\). In algebraic combinatorics, SNP has become a unifying polyhedral regularity condition for Schur-type functions, Schubert-type functions, Macdonald-type functions, chromatic symmetric functions, cluster variables, and related families; it is also closely tied to generalized permutahedra, discrete polymatroids, and integer decomposition phenomena [1703.02583].

## 1. Definition and polyhedral interpretation

For a polynomial
\[
f(x_1,\dots,x_n)=\sum_{\alpha\in\mathbb Z_{\ge0}^n} c_\alpha x^\alpha,
\]
its support is the set of exponent vectors with nonzero coefficient, and its Newton polytope is the convex hull of that support. The SNP condition requires that no lattice point of \(\Newton(f)\) be “missing” from the monomial support. Several papers also formulate the same condition for Laurent polynomials, replacing \(\mathbb Z_{\ge0}^n\) by \(\mathbb Z^n\); the geometric content is unchanged.

This condition is stronger than mere convexity of the exponent set and weaker than statements about coefficient size or positivity. It is purely about support. In the formulation used for supersymmetric Schur polynomials, SNP is equivalently the statement that the Newton polytope is the integer-hull of the support [2507.22528].

A second vocabulary comes from discrete convex analysis. Bechtloff Weising and Black define a finite set \(S\subset\mathbb Z^n\) to be \(M\)-convex if there is an integral generalized permutahedron \(P\subset\mathbb R^n\) such that \(S=P\cap\mathbb Z^n\); in that situation saturation is automatic. For homogeneous symmetric polynomials, Wang, Zhang, and Zhang record an equivalent formulation: \(M\)-convexity means that the polynomial is SNP and its Newton polytope is a generalized permutahedron [2508.00336].

## 2. Schur-theoretic foundations

The classical model for SNP is the Schur polynomial. Rado’s theorem identifies \(\Newton(s_\lambda)\) with the permutahedron \(P_\lambda\), and the Kostka expansion shows that the support is governed by dominance order. In the formulation used by Wang, Zhang, and Zhang,
\[
\mathrm{supp}(s_\lambda)=\{\mu\mid \mu\sqsubseteq\lambda\},
\]
so the Schur support already fills all lattice points of the \(\lambda\)-permutahedron. The 2017 survey of Monical, Tokcan, and Yong treats this as the basic prototype and records that skew Schur polynomials are SNP as well [2401.14632].

Several later families are proved SNP by comparison with the Schur case. For dual \(k\)-Schur polynomials \(S^{(k)}_\lambda\), Wang, Zhang, and Zhang prove support coincidence with the ordinary Schur polynomial indexed by the same \(k\)-bounded partition:
\[
\mathrm{supp}\bigl(S^{(k)}_\lambda(x)\bigr)=\{\mu\mid\mu\sqsubseteq\lambda\}=\mathrm{supp}\bigl(s_\lambda(x)\bigr),
\]
hence \(\Newton(S^{(k)}_\lambda)=\Newton(s_\lambda)\), and \(S^{(k)}_\lambda\) is SNP [2401.14632].

The same Schur-permutahedral structure persists in inhomogeneous settings. Escobar and Yong prove that every symmetric Grothendieck polynomial \(G_\lambda\) has SNP, that each homogeneous component \(G_\lambda^{(k)}\) has SNP, and that \(\Newton(G_\lambda^{(k)})\) is a permutahedron \(\Perm(\mu^{(k)})\), where \(\mu^{(k)}\) is the dominance-maximal partition of the relevant degree appearing in the Schur expansion [1705.07876].

## 3. Main proof paradigms

Several distinct proof mechanisms recur across the SNP literature.

| Method | Representative family | Output |
|---|---|---|
| Dominance order and permutahedra | Schur-type and chromatic symmetric functions | support equals lattice points of a permutahedron |
| Multidegrees and discrete polymatroids | double Schubert polynomials | support is the lattice set of a discrete polymatroid |
| Total unimodularity and Hoffman–Kruskal | supersymmetric Schur polynomials | integral polyhedron from hook inequalities |
| \(M\)-convex induction | non-symmetric Macdonald polynomials | support equals integer points of a generalized permutahedron |

The dominance-order method is the oldest. It uses explicit top partitions, Rado’s theorem, and containment of permutahedra. Nguyen–Giao–Hiep–Thuy extend this logic to a large class of “good” symmetric polynomials, proving that suitable Schur-linear combinations inherit both SNP and the integer decomposition property [2205.03903].

A second method interprets the polynomial as a multidegree. Castillo–Cid–Li–Montaño–Zhang’s framework, as applied by Fink, Mészáros, and St. Dizier, shows that the support of the multidegree polynomial of a Cohen–Macaulay prime ideal is a discrete polymatroid. Double Schubert polynomials arise as multidegrees of Schubert determinantal ideals; after sign-flip and standardization, the support is therefore a discrete polymatroid in \(\mathbb N^{2p}\), which yields SNP [2109.10299].

A third method is the discrete-convex approach. For non-symmetric Macdonald polynomials \(E_\mu\), Bechtloff Weising and Black prove that \(\supp(E_\mu)\) is \(M\)-convex by induction using the Knop–Sahi recurrence and a strong \(M\)-convexity lemma. Since an \(M\)-convex support is the lattice set of an integral generalized permutahedron, SNP follows immediately [2508.00336].

A fourth method, introduced by Dang and Nguyen, is polyhedral integrality via total unimodularity. They encode the support of a supersymmetric Schur polynomial by hook inequalities and a size constraint, construct a constraint matrix with a consecutive-ones structure after row-sign normalization, deduce total unimodularity from the Heller–Tompkins / Fulkerson–Gross criterion, and then apply the Hoffman–Kruskal criterion to prove integrality of the defining polyhedron. Dang and Nguyen state that, to their knowledge, this is the first application of total unimodularity to the SNP problem [2507.22528].

## 4. Representative modern families

For supersymmetric Schur polynomials, Dang and Nguyen work with two alphabets \(x=(x_1,\dots,x_k)\) and \(y=(y_1,\dots,y_\ell)\), and with partitions in the \((k,\ell)\)-hook. The tableau model of Berele–Regev gives
\[
S_\lambda(x,y)=\sum_{T\in SSYT_{k,\ell}(\lambda)} x^{a(T)}y^{b(T)}.
\]
Their main theorem states that for every partition \(\lambda\) in the \((k,\ell)\)-hook, \(S_\lambda(x,y)\) has SNP. The crucial support description is
\[
\Supp S_\lambda
=\{(a,b)\in\mathbb N^{k+\ell}:
\sum_{i=1}^r a_i\le \sum_{i=1}^r \lambda_i,\;
\sum_{j=1}^s b_j\le \sum_{j=1}^s \lambda'_j,\;
|a|+|b|=|\lambda|\},
\]
namely the hook inequalities plus the size constraint [2507.22528].

For chromatic symmetric functions, Matherne, Morales, and Selover prove an explicit permutahedral SNP statement. If \(d\) is a Dyck path and \(G(d)\) its indifference graph, then for every \(k\),
\[
\Newton(X_{G(d)}(x_1,\dots,x_k))=P_{\lambda^{gr}(d)}^{(k)},
\]
where \(\lambda^{gr}(d)\) is the partition given by the ordinary first-fit coloring of \(G(d)\). Thus the support is exactly the set of lattice points of that permutahedron. They extend the same conclusion to incomparability graphs of \((3+1)\)-free posets [2201.07333].

For non-symmetric Macdonald polynomials, Bechtloff Weising and Black prove that the supports are \(M\)-convex and therefore SNP. They also record a Bruhat-order description:
\[
E_\mu=\sum_{\lambda\le\mu} a_\lambda(q,t)\,x^\lambda,\qquad
\mathrm{supp}(E_\mu)=\{\lambda:\lambda\le\mu\},
\]
so the SNP statement becomes the equality
\[
\conv\{\lambda\le\mu\}\cap\mathbb Z^n=\{\lambda\le\mu\}.
\]
This resolves a 2019 conjecture of Monical–Tokcan–Yong [2508.00336].

For immanants of combinatorial matrices, Zhang proves SNP for all immanants of Giambelli matrices and verifies SNP for Jacobi–Trudi immanants in two special cases: \(\nu=(n)\) for any skew partition \(\lambda/\mu\), and \(\nu=(n-1,1)\) when \(\lambda/\mu\) is a border strip. The proofs use planar-network and lattice-path methods together with character evaluations on Young subgroups [2604.06615].

## 5. Discrete convexity, representation theory, and decomposition properties

SNP is frequently accompanied by stronger polyhedral properties. Nguyen–Giao–Hiep–Thuy prove that a “good” linear combination of Schur polynomials has both SNP and the integer decomposition property (IDP). Their proof uses Rado’s theorem, Schur saturation, and tableau decomposition, and yields a uniform source of IDP for Newton polytopes of symmetric polynomials [2205.03903].

A related direction is the use of SNP to prove IDP for specific lattice polytopes. Hong and Nasr study a 2-partition-maximal symmetric polytope \(P\subset\mathbb R^3\) lying in the hyperplane \(x_1+x_2+x_3=N\). For \(s=s_\lambda+s_\mu\) they define
\[
ts(x)=\sum_{i=0}^t s_{(t-i)\lambda+i\mu},
\]
prove that \(ts\) has saturated Newton polytope for every \(t\ge0\), and combine this with a general reduction to conclude that every 2-partition maximal symmetric polytope in a hyperplane of \(\mathbb R^3\) has IDP [2501.04191].

In representation theory, Besson, Jeralds, and Kiers study Demazure characters. For a dominant weight \(\lambda\) and \(w\in W\), the Demazure polytope is
\[
P_w(\lambda)=\Conv\{\mu:[e^\mu]\chi_w(\lambda)\neq0\}.
\]
They prove the vertex description
\[
P_w(\lambda)=\Conv\{u\cdot\lambda:u\le w\}
\]
and an inequality description via the Demazure product. In classical types \(A,B,C,D\) they show that every lattice point of \(P_w(\lambda)\) is realized as a weight, so Demazure characters are saturated; for \(G=GL_n\), this recovers the saturated-Newton-polytope property for key polynomials [2202.05405].

SNP also feeds into Ehrhart-theoretic questions. Bayer, Goeckner, Hong, McAllister, Olsen, Pinckney, Vega, and Yip prove that the Newton polytopes arising from Schur polynomials and inflated symmetric Grothendieck polynomials have IDP by using the fact that both families have SNP, and they give reflexivity classifications for these polytopes [2005.09628].

## 6. Failures, conjectures, and unresolved directions

SNP is widespread, but it is not automatic. Monical, Tokcan, and Yong record that the discriminant \(\Delta_n\) is SNP for \(n\le4\) but fails for \(n\ge5\), providing an explicit negative benchmark within algebraic combinatorics [1703.02583].

A major unresolved direction concerns Kronecker products. For
\[
s_\lambda*s_\mu=\sum_{\nu\vdash n} g(\lambda,\mu,\nu)\,s_\nu,
\]
the Monical–Tokcan–Yong conjecture predicts that the monomial expansion has SNP for all \(\lambda,\mu\vdash n\). González, Ikenmeyer, and Serrano prove this in two infinite families: \(\ell(\lambda)\le2\), \(\ell(\mu)\le3\), \(\mu_1\ge\lambda_1\); and \(\ell(\lambda)\le3\), \(\ell(\mu)\le2\) in the 3-variable specialization. Their proofs use Horn inequalities for positivity of Littlewood–Richardson coefficients and an analysis of the relevant integer polytopes [2311.10276].

Another partially open direction concerns immanants of Jacobi–Trudi matrices. Zhang verifies SNP in the permanent case and in the \((n-1,1)\) border-strip case, and states Conjecture 1.2 that \(\Imm_\nu H(\lambda,\mu)\) is SNP for all \(\nu\vdash n\) [2604.06615].

Cluster algebra examples show that finite-type behavior can be especially rigid. Mattoo and Sherman-Bennett prove that for cluster algebras of types \(A\) and \(D\), if the initial seed has principal coefficients or boundary frozen variables, then all cluster variable Newton polytopes are saturated; in type \(A\) they are in fact empty. They also note that for most other surfaces one finds explicit counterexamples [2012.07500].

Several SNP families are now linked to Lorentzian phenomena. Matherne, Morales, and Selover conjecture that the relevant chromatic symmetric functions are Lorentzian and prove this for abelian Dyck paths, while Dang and Nguyen conjecture that each supersymmetric Schur polynomial \(S_\lambda\) is Lorentzian in the sense of Brändén–Huh [2201.07333]. A plausible implication is that SNP increasingly functions not as an isolated support property, but as one component of a broader discrete-convex package involving generalized permutahedra, \(M\)-convexity, log-concavity, and polyhedral integrality.

Source: https://www.emergentmind.com/topics/saturated-newton-polytope-property