REVIEW 4 major objections 4 minor 18 references
Invertible sheaves and $\Pi$-invertible sheaves on the isomeric supergrassmannian and its toric subvarieties
T0 review · 4 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash
Pith's one-line read The paper proves that the Picard group and Pi-Picard set of the isomeric supergrassmannian QGr(r,n) vanish except on factors of Pi-projective space, and that for toric orbit closures they are computed by counting simplex factors in the…
desk verdict Useful combinatorial classification of Pic and Pi-Picard on type-Q supergrassmannians and orbit closures, but the toric part rests on a lemma whose proof is too quick. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The engine is the momentum-map polytope $P$ attached to each $Q(1)^n$-orbit closure, a subpolytope of the hypersimplex whose edges point along coordinate or coordinate-difference directions and whose normal fan is the decorated fan of the toric supervariety. Invertible sheaves are encoded by integer-valued transition functions on the 1-skeleton of $P$, and the cocycle condition forces every transition function to vanish except along edges belonging to one of the $pc(P)$ parallel-component directions, which are exactly the $\mathbb{P}^1_\Pi$ factors. For $\Pi$-invertible sheaves the additional continuous data $c\in\mathbb{C}$ appear, and the key dichotomy (Lemma 4.11) asserts that a sheaf nontrivial on a maximal simplicial face forces all incident edges to lie in square 2-faces, so $P$ is a product of that simplex with another polytope; thus nonzero $\Pi$-invertible sheaves live only on simplex factors $\mathbb{P}^d_\Pi$, where the relation $c^2=\ell$ comes from comparing transition functions around a triangle. The sheaves $\mathcal{O}_\Pi(\ell)$ are then realized as symmetric powers of the tautological sheaf and its dual, with a rescaled odd involution.
What would settle it
Find a $Q(1)^n$-orbit closure inside some $\operatorname{QGr}(r,n)$ whose defining polytope is not a product with a simplex but whose $\Pi$-Picard set is nonzero; the paper's Lemma 4.11 says no such orbit closure exists. Concretely, enumerate subpolytopes of the hypersimplex $\Delta_{r,n}$ with the allowed edge directions, test the transition-function equations for a pentagon-type configuration with a nonzero $(\ell,c)$, and check that the configuration is realized by an actual orbit closure.
Extended reading notes
Core claim
On its own terms, the paper's central result is Theorem 1. Let $X$ be $\operatorname{QGr}(r,n)$ or the closure of a $Q(1)^n$ orbit. Then $\operatorname{Pic}(X)$ is the product of one copy of $\mathbb{Z}$ for each factor $\mathbb{P}^1_\Pi$ of $X$, and $\operatorname{Pic}^\Pi(X)$ is the pointed set of tuples $(\ell_i,c_i)$ in a product of copies of $\mathbb{Z}\oplus\mathbb{C}$, one for each factor $\mathbb{P}^{d_i}_\Pi$, with at most one $c_i\neq 0$ and, when $d_i>1$, the constraint $c_i^2=\ell_i$. In particular, for the supergrassmannian itself, nonzero classes occur only when $\operatorname{QGr}(r,n)$ is a $\Pi$-projective space: $\operatorname{Pic}(\mathbb{P}^1_\Pi)=\mathbb{Z}$ and $\operatorname{Pic}^\Pi(\mathbb{P}^1_\Pi)=\mathbb{Z}\oplus\mathbb{C}$, while for $\mathbb{P}^{n-1}_\Pi$ with $n>2$ the $\Pi$-Picard set is $\{(\ell,c): c^2=\ell\}$; all other $\operatorname{QGr}(r,n)$ have trivial Picard group and trivial $\Pi$-Picard set. The paper reads this as saying that the presence of nontrivial invertible sheaves depends entirely on factors of $\Pi$-projective space, and constructs the nonzero sheaves as symmetric powers $\mathcal{O}_\Pi(\ell)$ of the tautological sheaf and its dual, with the odd involution rescaled appropriately.
Load-bearing premise
The whole toric classification rests on the geometric dichotomy of Lemma 4.11: a $\Pi$-invertible sheaf that is nontrivial on a maximal simplicial face of the defining polytope must force every other edge incident to that face to lie in a square two-dimensional face, so the polytope splits as a product of that face with another polytope; if that dichotomy fails, the stated $\Pi$-Picard sets for orbit closures are incomplete.
Editorial extensions
If this is right
- For $\operatorname{QGr}(r,n)$ itself, nontrivial invertible or $\Pi$-invertible sheaves exist only when the variety is a $\Pi$-projective space; otherwise both $\operatorname{Pic}$ and $\operatorname{Pic}^\Pi$ vanish.
- For a supertorus orbit closure, $\operatorname{Pic}(X)\cong\mathbb{Z}^{pc(P)}$, where $pc(P)$ is the number of independent parallel-edge families of the defining polytope, equivalently the number of $\mathbb{P}^1_\Pi$ factors.
- The $\Pi$-Picard set of an orbit closure is supported on a single simplex factor: elements are tuples $(\ell_i,c_i)$ with at most one $c_i\neq 0$, subject to $c_i^2=\ell_i$ on factors of dimension greater than 1.
- Every nonzero sheaf is explicit: invertible sheaves are tensor products of $\mathcal{O}(\ell)$ on $\mathbb{P}^1_\Pi$ factors, and $\Pi$-invertible sheaves are such products tensored with $\mathcal{O}_\Pi(\ell)$ on one higher-dimensional factor, realized as symmetric powers of the tautological sheaf or its dual.
- Because $\Pi$-invertible sheaves characterize morphisms to $\Pi$-projective superspace, the polytope data also control which of these supervarieties admit such maps.
Reading between the lines
- A purely combinatorial strengthening of Lemma 4.11 seems within reach: for hypersimplex subpolytopes, $\operatorname{Pic}^\Pi(X)=0$ unless the polytope factors as a simplex times a polytope whose remaining edges point in coordinate directions; this could be verified by enumerating subpolytopes of $\Delta_{r,n}$.
- The same transition-function analysis applies to toric supervarieties beyond orbit closures, where the $\Pi$-Picard set need not be a finite set: the paper's pentagon example shows it can be a surface defined by quadratic equations.
- Since $\mathcal{O}_\Pi(\ell)$ is constructed as a symmetric power with a rescaled odd involution, the paper implicitly supplies the odd transition functions needed to build explicit embeddings of these toric supervarieties into $\Pi$-projective spaces.
- The appearance of $\mathbb{C}\cong H^1(X,\Omega^1)$ as the continuous parameter suggests arithmetic refinements: over other base fields the condition $c^2=\ell$ may interact with the field's square classes, a direction the paper leaves open.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper studies invertible sheaves and Pi-invertible sheaves on the isomeric supergrassmannian QGr(r,n) and on closures of Q(1)^n orbits in it. Theorem 1 asserts that Pic(X) is a free abelian group whose rank equals the number of P^1_Pi factors of X, and that Pi-Pic(X) is the pointed set of tuples (ell_i,c_i) with at most one nonzero c_i and satisfying c_i^2 = ell_i for factors of dimension greater than one. For QGr(r,n) itself, this gives Pic(QGr(r,n)) = Z for (r,n)=(1,2) and 0 otherwise, and Pi-Pic(QGr(r,n)) = Z+C for (1,2), {(ell,c): c^2=ell} for r=1 or n-1 with n>2, and 0 otherwise. The proof proceeds by computing Pic and Pi-Pic of P^n_Pi, using standard extensions to reduce the general QGr case, and then using toric geometry and a combinatorial analysis of hypersimplex subpolytopes to handle orbit closures.
Significance. If correct, the paper gives a clean and largely elementary computation of Picard groups and Pi-Picard sets for a family of supervarieties for which complete written proofs were previously missing, and it provides a combinatorial criterion (the parallel component number) that makes the computation for toric subvarieties tractable. The paper is clearly written and contains explicit toric chart computations and transition functions; it also correctly identifies the exceptional cases. The reliance on the author's prior work [7],[8] for the foundations of toric supervarieties and on [13],[14] for structural facts about Pi-invertible sheaves is reasonable and does not constitute circularity. The main weakness is that several load-bearing steps, especially the polytopal dichotomy in Lemma 4.11, are asserted rather than proved.
major comments (4)
- [Section 4.2, Lemma 4.11] The proof of Lemma 4.11 is the load-bearing step for the toric Pi-Picard classification, but the product-or-trivial dichotomy is not demonstrated. The proof asserts that if a Pi-invertible sheaf is nontrivial on a maximal simplicial face S, then Lemma 4.10 forces all other edges incident to vertices of S to lie in square 2-faces, and concludes that P is a product of S with another polytope. Neither implication follows from the cited lemmas: Lemma 4.10 only handles two triangular faces meeting along an edge and not part of a larger simplicial face, and it says nothing about edges that meet S only at a vertex; moreover, even if each additional edge lies in a square, a collection of squares around a simplex does not force a global product structure, and configurations such as a square pyramid or a chain of adjacent simplicial faces are not ruled out. Since Proposition 4.12 and the toric part of Theorem 1 depend on this lemma, the proof needs a genuine polytopal argument using the restriction that edges of P are parallel to x_i^* or x_i^*-x_j^* within a hypersimplex subpolytope.
- [Section 3.1, Lemma 3.3] The computation of Pic(P^2_Pi)=0, which is the base case for the vanishing of Pic(QGr(r,n)) for 0<r<n-1, is not actually carried out. After writing the even coordinate rings of the affine charts and the general form of the transition functions, the proof ends with 'An explicit computation reveals that this is not possible unless ell=0' without exhibiting the contradiction. Since this lemma is used to reduce the general case via standard extensions, the missing computation should be supplied or at least summarized in enough detail that the reader can verify the claim.
- [Section 4.2, Lemma 4.10 and Section 4.1, Lemma 4.5] The proof of Lemma 4.10 begins with 'Without loss of generality, two such triangles occur at the vertices 1100, 1010, 0110, and 0011.' This reduction is not justified: it assumes that any pair of triangular faces meeting along an edge in a subpolytope of a hypersimplex is equivalent under symmetries preserving the decorated fan and the sign conventions of Figure 3. The subsequent computation, which shows that the two triangles impose the incompatible conditions c^2=ell and c^2=-ell, depends entirely on this normalization. If a configuration exists in which the two triangles induce the same orientation, Lemma 4.10 would be false. The same sign issue underlies Lemma 4.5, where the assertion that adjacent triangles in QGr(2,4) have disagreeing clockwise directions is stated without proof. The WLOG claim and the sign convention need to be made precise.
- [Section 3.2, Proposition 3.8 and Section 4.2, Lemma 4.8] Two structural statements are asserted without adequate proof. In Proposition 3.8, the passage from the cocycle condition 'every cycle containing e contains another edge parallel to e' to the parallel component number pc(P) of Definition 3.6 is not demonstrated, and the equivalence between this cycle condition and the disconnection property of the 1-skeleton deserves a graph-theoretic argument. In Lemma 4.8, the isomorphism PicPi(X/J^2) ≅ (Z⊕C)^{rk Pic(X_red)} is stated without proof; since this underlies the description of PicPi(X/J^2) in the toric case, it should be justified, for instance by computing H^1(Omega^1) for complete toric varieties.
minor comments (4)
- [References] The arXiv number for reference [8] is given as 'math/0507198', which appears to be a typo; the linked URL points to arXiv:2502.15977.
- [Section 4.2, Proposition 4.12] The proof that at most one c_i can be nonzero says that 'f1 and f2 commute' after forming f1 f2 f1^{-1} f2^{-1}=1. Since transition functions are elements of O^×_X, which in the super setting are not necessarily commutative, this step is confusing; the intended noncommutativity of even and odd units should be explained explicitly.
- [Definition 3.6] Part (a) of Definition 3.6 is ambiguous: 'the 1-skeleton becomes disconnected upon removal of the edges parallel to L' could be read as a property of a single removal rather than of all removals of edges in that direction. Please rephrase to make clear that one removes all edges parallel to L and checks that the number of connected components increases.
- [Section 3.2, Lemma 3.5] The injectivity of Pic(X) → Pic(X_red) for toric varieties uses not only H^1(X, Omega^q)=0 for q>1 but also H^1(X, O_X)=0; this latter vanishing should be stated explicitly.
Circularity Check
No circularity: Picard and Pi-Picard classifications are derived from explicit transition-function computations; self-citations supply framework, not the theorem, and the acknowledged Lemma 4.11 gap is a rigor issue, not a circular reduction.
full rationale
The paper's central claims are proved self-containedly rather than read off from its inputs. Pic(QGr(r,n)) is computed from a filtration/injectivity argument (Lemma 3.2) and direct chart computations for P^2_Pi (Lemma 3.3); the toric Picard result (Proposition 3.8) is proved by cocycle computations on the 1-skeleton, with pc(P) treated combinatorially via Definition 3.6(a). The Pi-Picard results are likewise computed directly: Lemma 4.3 derives the c^2 = ell condition from transition functions on P^n_Pi, Lemma 4.5 computes PicPi(QGr(2,4)) from adjacent triangle embeddings, and Lemma 4.10 repeats that local computation for toric cells. Citations to the author's [7] and [8] set up the decorated-fan and toric-supervariety language, and [13] and [14] supply standard structural facts such as J_X/J_X^2 being isomorphic to Omega^1; none of these citations imports the Picard or Pi-Picard classification being proved. There is a genuine proof gap in Lemma 4.11, where the claim that local square incidences force P to be a product is asserted rather than proved, and Example 4.14 concedes that the analogous statement fails for arbitrary decorated fans; this is a correctness or rigor risk, not a circular reduction of the theorem to its inputs. No equation in the paper is equivalent by construction to the result it is used to prove.
Assumptions & free parameters
assumptions (6)
- domain assumption J_X^{2i}/J_X^{2i+1} is isomorphic to Omega^{2i}_{X_red} for QGr(r,n) and its toric subvarieties.
- standard math The Hodge numbers of the classical Grassmannian satisfy h^{p,q}=0 unless p=q.
- standard math For a toric variety X_red, H^1(X_red, Omega^q)=0 for q>1 (Danilov's theorem).
- domain assumption The closure of a Q(1)^n orbit in QGr(r,n) is a normal toric supervariety whose decorated fan and polytope P is a subpolytope of Delta_{r,n} with edges parallel to x_i^* or x_i^*-x_j^*, and with fermionic sheaf F_X isomorphic to Omega^1_{X_red}.
- domain assumption Pi-invertible sheaves are classified by H^1(X, O_X^times), and on P^n_Pi the transition functions for X/J^2 have the form t_i^ell(1+c xi_i).
- domain assumption Every 2-face of a subpolytope P of Delta_{r,n} arising from an orbit closure is a triangle or a square.
Cite this review
Pith. "Pith review of Invertible sheaves and $\Pi$-invertible sheaves on the isomeric supergrassmannian and its toric subvarieties." pith.science (2026). https://pith.science/paper/UZGM3UZY
@misc{pith2026250622386,
author = {Pith},
title = {Pith review of: Invertible sheaves and $\Pi$-invertible sheaves on the isomeric supergrassmannian and its toric subvarieties},
year = {2026},
howpublished = {\url{https://pith.science/paper/UZGM3UZY}},
note = {Machine review of arXiv:2506.22386}
}
abstract
We provide an elementary proof that with the exceptions of certain $\Pi$-projective spaces, both the Picard group and the $\Pi$-Picard set of the isomeric (i.e. type-Q) supergrassmannian are trivial. We extend this technique to show that the Picard group and the $\Pi$-Picard set of a supertorus orbit closure within the isomeric supergrassmannian can be easily calculated from its defining polytope by counting the number of simplex factors. Since the presence of nontrivial invertible sheaves and $\Pi$-invertible sheaves depends entirely on factors of $\Pi$-projective space, we construct them as symmetric powers of the tautological sheaf and its dual.
Figures
Figures from the paper (1 more)
Reference graph
Works this paper leans on
-
[7]
Jankowski, Toric supervarieties with one odd dimension , Transform
E. Jankowski, Toric supervarieties with one odd dimension , Transform. Groups, (2024). https: //link.springer.com/article/10.1007/s00031-024-09889-6
-
[8]
Jankowski, Reduced superschemes and the combinatorics of toric supervarieties
E. Jankowski, Reduced superschemes and the combinatorics of toric supervarieties. arXiv preprint, 2025, math/0507198. https://arxiv.org/abs/2502.15977
arXiv 2025
-
[13]
Y. I. Manin, I. B. Penkov, and A. A. Voronov , Elements of supergeometry, J. Sov. Math., 51 (1990), pp. 2069–2083. https://link.springer.com/article/10.1007/BF01098184
-
[14]
Noja , Supergeometry of Π-projective spaces, J
S. Noja , Supergeometry of Π-projective spaces, J. Geom. Phys., 124 (2018), pp. 286–299. https: //www.sciencedirect.com/science/article/pii/S0393044017302905
work page 2018
- [1]
-
[2]
S. L. Cacciatori and S. Noja , Projective superspaces in practice, J. Geom. Phys., 130 (2018), pp. 40–62. https://www.sciencedirect.com/science/article/pii/S0393044018301955
work page 2018
-
[3]
D. Cox, J. Little, and H. Schenck , Toric Varieties, vol. 124 of Graduate Studies in Mathe- matics, American Mathematical Society, 2011
work page 2011
-
[4]
Danilov, The geometry of toric varieties , Russ
V. Danilov, The geometry of toric varieties , Russ. Math. Surv., 33 (1978), pp. 97–154. https: //iopscience.iop.org/article/10.1070/RM1978v033n02ABEH002305
Show all 18 references
-
[5]
Ewald, Algebraic Geometry, vol
G. Ewald, Algebraic Geometry, vol. 168 of Graduate Texts in Mathematics, Springer-Verlag, 1996
1996
-
[6]
Gel’fand, R
I. Gel’fand, R. M. Goresky, R. D. MacPherson, and V. V. Serganova , Combinatorial geometries, convex polyhedra, and schubert cells , Adv. Math., 63 (1987), pp. 301–316. https:// www.sciencedirect.com/science/article/pii/0001870887900594
1987
-
[9]
Kwok, The geometry of Π-invertible sheaves, J
S. Kwok, The geometry of Π-invertible sheaves, J. Geom. Phys., 86 (2014), pp. 134–148. https: //www.sciencedirect.com/science/article/pii/S0393044014001557
2014
-
[10]
LeBrun, Y.-S
C. LeBrun, Y.-S. Poon, and J. R. O. ˜Wells, Projective embeddings of complex su- permanifolds, Commun. Math. Phys., 126 (1990), pp. 433–452. https://projecteuclid. org/journals/communications-in-mathematical-physics/volume-126/issue-3/ Projective-embeddings-of-complex-superman...
1990
-
[11]
Y. I. Manin , Gauge Field Theory and Complex Geometry , vol. 289 of Grundlehren der Mathema- tischen Wissenschaften, Springer-Verlag, 1988
1988
-
[12]
, Topics in Non-Commutative Geometry , Porter Lectures, Princeton University Press, 1991
1991
-
[15]
I. B. Penkov , Borel-weil-bott theory of classical lie superalgebras , J. Sov. Math., 51 (1990), pp. 2108–2140. https://link.springer.com/article/10.1007/BF01098186
1990 doi
-
[16]
I. B. Penkov and I. A. Skornyakov , Projectivity and D-affinity of flag supermanifolds , Russ. Math. Surv., 40 (1985), pp. 233–234. https://iopscience.iop.org/article/10.1070/ RM1985v040n01ABEH003546
1985
-
[17]
I. B. Penkov and A. S. Tikhomirov, Linear ind-grassmannians, Pure Appl. Math. Q., 10 (2014). https://math.constructor.university/penkov/papers/Ind-grassmanians.pdf
2014
-
[18]
Polishchuk , A0|1-torsors, quotients by free A0|1-actions, and embeddings into Π-projective spaces and super-grassmannians G(1|1, n|n), Commun
A. Polishchuk , A0|1-torsors, quotients by free A0|1-actions, and embeddings into Π-projective spaces and super-grassmannians G(1|1, n|n), Commun. Math. Phys., 402 (2023), pp. 2011–2029. https://par.nsf.gov/biblio/10521212
2023
Reviewed August 6, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.