REVIEW 2 major objections 5 minor 1 cited by
Cohomology on the incidence correspondence and related questions
T0 review · 2 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash
Pith's one-line read This paper proves a terminating recursive formula for the cohomology characters of line bundles on the incidence correspondence in positive characteristic, solves the recursion explicitly in characteristic 2 with Nim polynomials, and…
desk verdict A serious, technically strong paper that gives the first full cohomology description for incidence correspondences in all positive characteristics; the main soft spot is the reliance on an unpublished resolution from [RV23] in Theorem 3.2(3). 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 load-bearing object is the family $F^d_r$ of vector bundles on $\mathbb P^1$ defined as the kernel of the map $D^dU\otimes\mathcal O_{\mathbb P^1}\to D^{d-r}U\otimes\mathcal O_{\mathbb P^1}(r)$; its dual is the bundle of principal parts of order $r-1$ of $\mathcal O_{\mathbb P^1}(d)$, whose fibers record Taylor expansions of sections up to that order. Theorem 3.2 gives a recursive, torus-equivariant description of the splitting type of $F^d_r$ into line bundles in terms of smaller $F$'s and Frobenius pullbacks. That splitting recursion is transported to projective space through long exact sequences: the cohomology of $F^d_r(e)$ filters the cohomology of $D^dR(e)$, and the failure of exactness is encoded in the graded Han–Monsky representation ring of the modules $\delta_a=k[T]/(T^a)$. In characteristic $2$ this ring multiplication is governed by Nim sums, which yields the closed formulas for cohomology and for the Weak Lefschetz Property.
What would settle it
For $p=2$, $n=11$, $d=6$, $e=5$, compute directly the cokernel of multiplication by $\omega=x_1y_1+\cdots+x_{11}y_{11}$ on the bi-graded module $M_{6,5}$ and compare its dimension with the truncated Schur sum in Corollary 7.2; the theorem predicts an extra elementary-symmetric term invisible for $n\le10$, so a mismatch would disprove Theorem 1.3. For odd characteristics, do the same direct computation for $p=3$, $d=3$, $e=2$, where the paper predicts a 51-dimensional $H^1(\mathbb P^4,D^3R(2))$.
Extended reading notes
Core claim
Over a field of characteristic $p>0$, the paper establishes the identity $$ h^i(D^dR(e))=\sum_{a,b} \Phi_{d-ap,e-bp}\cdot F_p\big(h^i(D^aR(b))\big) $$ for $e\ge d-1$ and $i=0,1$, where $D^dR(e)$ is the divided power of the tautological subsheaf twisted by $\mathcal O(e)$, $\Phi$ is built from products of $p$-truncated complete symmetric polynomials, and $F_p$ is the Frobenius action on the character ring. The sum ranges over $0\le a\le d/p$ and $-1\le b\le (d+e)/p$, and it terminates because $h^i(D^aR(b))$ is the classical Schur-polynomial character whenever $a<p$. In characteristic $2$, Corollary 7.2 solves the recursion into an explicit sum of Frobenius twists of Nim symmetric polynomials times truncated Schur polynomials, correcting the earlier conjecture by adding terms that first appear when the number of variables is at least $11$. The same machinery yields Theorem 8.1, an if-and-only-if criterion for the Weak Lefschetz Property of Artinian monomial complete intersections in characteristic $2$.
Load-bearing premise
The load-bearing premise is Lemma 3.7: certain binomial coefficients $\binom{f}{v p^{e-1}}$ with $0\le v<p$ and $v p^{e-1}\le f<p^e$ do not vanish in $k$, and this is what makes the leading coefficients of the key locally split inclusion units; if any such coefficient vanished, the recursive decomposition of $F^d_r$ and the cohomology recursion would collapse.
Editorial extensions
If this is right
- All cohomology characters $h^i(D^dR(e))$ in any characteristic $p$ can be computed by a terminating recursion starting from the known $d<p$ base case.
- In characteristic $2$ the explicit formula with Nim polynomials and truncated Schur polynomials gives a closed-form computation and repairs a previously stated conjecture by adding terms that only occur with at least $11$ variables.
- Theorem 8.1 supplies a finite, checkable criterion for the Weak Lefschetz Property of monomial complete intersections in characteristic $2$.
- Theorem 1.5 characterizes all socle degrees for which every monomial complete intersection satisfies WLP, recovering an earlier result and proving a conjecture from the literature on Lefschetz properties.
- The graded Han–Monsky ring obeys the degree formula $c+2j=a_1+\cdots+a_n-(n-1)$ for summands $\delta_c(-j)$ with $p\nmid c$, and in characteristic $2$ the presence of summands is decided by Nim sums.
Reading between the lines
- Beyond the paper: the mechanism should extend to other partial flag varieties of Picard rank 2 whose cohomology is controlled by the splitting of principal parts on $\mathbb P^1$.
- Beyond the paper: the correction to the characteristic-2 conjecture is a caution that computational checks in few variables can miss cohomology terms; for the incidence correspondence, new terms can first appear when the number of variables reaches $d+e+1$.
- Beyond the paper: the Han–Monsky summand rules suggest that a parallel analysis should yield strong Lefschetz characterizations in positive characteristic, since the Strong Lefschetz Property is encoded by the same summand shifts.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper develops recursive formulas for the sheaf cohomology of line bundles on the incidence correspondence over fields of positive characteristic. The main result, Theorem 1.1, expresses h^i(D^d R(e)) for e ≥ d−1 as a finite sum of terms Φ_{d−ap,e−bp}·F_p(h^i(D^a R(b))), with the characteristic-zero Borel–Weil–Bott result serving as the base case. The technical engine is Theorem 3.2, a recursive description of the T-equivariant splitting type of the sheaves F^d_r on P^1, which are duals of bundles of principal parts; the proof uses explicit local bases, differential operators, and Lucas' theorem. The paper then derives a non-recursive characteristic-2 formula in terms of Nim polynomials and truncated Schur polynomials (Theorem 1.3, correcting a conjecture of GRV24), relates the recursion to the graded Han–Monsky representation ring (Theorem 1.4), and gives a Weak Lefschetz Property criterion for monomial complete intersections in characteristic 2 (Theorem 8.1), together with a characterization of socle degrees that guarantee WLP (Theorem 1.5).
Significance. If the main results are correct, Theorem 1.1 provides the first infinite family of Picard-rank-2 flag varieties whose line-bundle cohomology is computed in every characteristic, and Theorem 1.3 gives a genuinely non-recursive description in characteristic 2. The paper is unusually concrete: the recursion is illustrated by explicit examples (Example 1.2), the base case is the classical Borel–Weil–Bott statement, no parameters are fitted, and several falsifiable numerical predictions are given. The splitting theorem for principal parts is a substantial technical contribution, and the connection between cohomology, Han–Monsky multiplication, and Lefschetz properties is elegant. The main weakness is that part (3) of Theorem 3.2 relies on an 'easy extension' of an unpublished preprint [RV23] without proof; this is the only load-bearing gap that I identified, together with a boundary-case omission in the proof of Theorem 4.2(5).
major comments (2)
- [§3.2, proof of Theorem 3.2(3)] The four-term resolution G• is asserted as 'an easy extension of the argument in [RV23, Section 4.5]' and is then used to identify F^{d−r}_{q−r}(q) with a cokernel and to prove the exactness of the first column in the displayed diagram. This resolution is load-bearing: case (3) is needed for Corollary 3.5 and for the proof of part (4), and hence for the main recursion in Theorem 1.1. Since [RV23] is an unpublished preprint and the resolution is not stated or proved in this manuscript, the central technical theorem is not independently verifiable as written. I request either a full proof of the resolution or a precise statement with enough detail for a reader to check it.
- [§4, proof of Theorem 4.2(5)] The displayed calculation proving part (5) applies Theorem 3.2(4) to the sheaves F^d_{r+1} and F^{d−1}_r, but this requires the hypotheses r+1 ≤ d and aq' ≤ r+1 < (a+1)q' for the same parameter a. When r = d, the first inequality fails; when r = (a+1)q'−1, the second inequality fails for the chosen a, and r+1 may even equal q, in which case F^d_{r+1} falls under case (1) rather than case (4) of Theorem 3.2. The proof does not explain how these boundary cases are handled or why the displayed identity remains valid. Since Theorem 4.2(5) is used for Corollary 4.4 and Proposition 5.4, and hence for the main cohomology recursion, this omission should be addressed explicitly.
minor comments (5)
- [Example 1.2] There is a typographical error in 'chararacteris-tics' on page 3; it should read 'characteristics'.
- [§3.2, proof of Theorem 3.2(2)] The statement 'It follows from (3.4) that the determinant of the complex is trivial' is correct but terse; adding a one-line degree and T-weight computation for the determinant of the complex would improve readability.
- [Notation, Theorem 3.2] The notation F^q(H_{a,0}) is used in Theorem 3.2 before its concrete interpretation as a span of differential operators in Corollary 3.8; a forward reference or a short explanation at first use would help the reader.
- [§8.2, Theorem 8.1] The function θ_q takes the value −∞ at r = 2q−1; in the proof of Theorem 8.1 the inequalities involving sums of θ_q values should be interpreted in the extended real sense. This is harmless but could be stated explicitly.
- [Introduction] The acronym SLP is introduced in the introduction but the paper focuses on WLP; consider either removing the SLP remark or adding a brief cross-reference to the cited literature.
Circularity Check
No circularity: the cohomology recursion has an independent characteristic-zero base case and no fitted parameters; the main dependency on [RV23] is an external technical lemma, not the target result.
full rationale
The paper's central recursion (Theorem 1.1) expresses h^i(D^dR(e)) as a finite sum of terms involving the same character with strictly smaller d, together with p-truncated symmetric polynomials, and the base case d < p is supplied by the independent Borel-Weil-Bott description in (1.2). The splitting theorem 3.2 is proved by induction on (d, r), with the four cases covering disjoint ranges and the inductive hypotheses used explicitly; Lemma 3.7, which guarantees the leading coefficients in Corollary 3.8 are units, is proved directly by Lucas' theorem. The Han-Monsky results (Theorem 5.2, Proposition 5.4, Corollary 5.6) are derived from the splitting theorem and the ideals I(d,r) and J(d,r), not assumed. The characteristic-2 non-recursive formula (Theorem 1.3 / Corollary 7.2) and the WLP criterion (Theorem 8.1) are algebraic consequences of the same recursion, with no fitted parameters. The one external load-bearing input is the resolution from [RV23, Section 4.5] used in the proof of Theorem 3.2(3); it is a self-citation to an unpublished preprint and the proof is omitted, so it is a correctness and completeness risk, but it is not circular: the resolution is stated as a general lemma about short exact sequences of locally free sheaves and does not assume the cohomology characters or the splitting type that the paper derives. No step defines its conclusion in terms of itself, and no fitted input is renamed as a prediction.
Assumptions & free parameters
assumptions (5)
- standard math Borel-Weil-Bott theorem over characteristic zero
- standard math Grothendieck's theorem on splitting of vector bundles on P1
- standard math Kumar's theorem on T-equivariant splitting
- standard math Lucas' theorem on binomial coefficients modulo p
- standard math Stanley's theorem on Lefschetz properties in characteristic zero
Cite this review
Pith. "Pith review of Cohomology on the incidence correspondence and related questions." pith.science (2026). https://pith.science/paper/55YH4L6R
@misc{pith2026241113450,
author = {Pith},
title = {Pith review of: Cohomology on the incidence correspondence and related questions},
year = {2026},
howpublished = {\url{https://pith.science/paper/55YH4L6R}},
note = {Machine review of arXiv:2411.13450}
}
read the original abstract
We study a variety of questions centered around the computation of cohomology of line bundles on the incidence correspondence (the partial flag variety parametrizing pairs consisting of a point in projective space and a hyperplane containing it). Over a field of characteristic zero, this problem is resolved by the Borel-Weil-Bott theorem. In positive characteristic, we give recursive formulas for cohomology, generalizing work of Donkin and Liu in the case of the 3-dimensional flag variety. In characteristic 2, we provide non-recursive formulas describing the cohomology characters in terms of truncated Schur polynomials and Nim symmetric polynomials. The main technical ingredient in our work is the recursive description of the splitting type of vector bundles of principal parts on the projective line. We also discuss properties of the structure constants in the graded Han-Monsky representation ring, and explain how our cohomology calculation characterizes the Weak Lefschetz Property for Artinian monomial complete intersections.
Forward citations
Cited by 1 Pith paper
-
The Gr\"obner basis for powers of a general linear form in a monomial complete intersection
An explicit description of the reduced Gröbner bases of (x1^{m1},...,xn^{mn},(x1+...+xn)^k) is obtained using lattice paths and a reflection map.
Reference graph
Works this paper leans on
-
[1]
[And81] Henning Haahr Andersen, On the structure of the cohomology of line bundles on G/B, J. Algebra 71 (1981), no. 1, 245–258. [And23] , Representation theory via cohomology of line bundles , Transform. Groups 28 (2023), no. 3, 1033–1058. [AK12] Henning Haahr Andersen and Masaharu Kaneda, Cohomology of line bundles on the flag variety for type G2, J. Pur...
work page 1981
-
[11]
[Liu24] Linyuan Liu, Cohomologie des fibr´ es en droites sur SL3/B en caract´ eristique positive: deux filtrations et cons´ equences, J. Eur. Math. Soc. (JEMS) 26 (2024), no. 4, 1365–1422 (French, with English and French su mmaries). [LN19] Samuel Lundqvist and Lisa Nicklasson, On the structure of monomial complete intersections in posi tive characteristic,...
work page Pith review arXiv 2024
-
[25]
[LP21] Linyuan Liu and Patrick Polo, On the cohomology of line bundles over certain flag schemes II , J. Combin. Theory Ser. A 178 (2021), Paper No. 105352,
work page 2021
Reviewed August 12, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.