Searcharxiv⌕ Search

arXiv subjects

Cristian Anghel

Publications and source records attributed to Cristian Anghel.

At least 19 recordsLinked to original sources

Boij-S"oderberg and Harder-Narasimhan polygons of vector bundles on projective space

Eisenbud and Schreyer proved that the cohomology table of a vector bundle $F$ on $\mathbb{P}^n$ is a unique positive combination of supernatural tables along a chain of root sequences. The slopes of the pieces, weighted by their ranks, form a concave polygon $Ψ_F$, the Boij-Söderberg polygon, with the same endpoints as the Harder-Narasimhan polygon $P_F$. We prove that $P_F\leqΨ_F$, in any characteristic, by testing subsheaves against sheaves with pure resolutions on thickened lines; the bound is optimal among bounds depending only on the ray of the table. The chain majorizes every other supernatural decomposition, and $Ψ$ is monotone under cancellations, hence upper semicontinuous in flat families. The comparison extends to coherent sheaves and, by linear projection, to sheaves of positive dimension. On $\mathbb{P}^2$ the defect is a sum of four cohomological terms; we characterize equality and show that the inequality survives the large volume limit of Bridgeland stability. For a smooth projective variety $X$ of dimension $n\geq2$ in characteristic zero, with $H$ ample and base-point-free, the mean normalized squared spread $G(\mathcal{E})$ of the roots in the cohomology table of every vector bundle $\mathcal{E}$ satisfies $G(\mathcal{E})\geq(K_X^2-2c_2(X))\cdot H^{n-2}/((n+1)H^n)$, a bound independent of $\mathcal{E}$. Since $G(\mathcal{E})=1$ exactly for arithmetically Cohen-Macaulay bundles, when this bound exceeds $1$ it is a uniform form of their absence.

math.AG↗

Generic splitting types and Terao's conjecture for seventeen and eighteen lines

We prove that the generic splitting type of the logarithmic bundle of an arrangement of at most eighteen lines in the complex projective plane is determined by the intersection lattice. In particular, Terao's freeness conjecture holds for seventeen and eighteen lines. For any number of lines, it holds for free arrangements whose smaller exponent is at most eight. By the theorem for at most sixteen lines, the problem reduces to a uniformity condition for lattices of seventeen lines: if all realizations have minimal Jacobian-syzygy degree at least seven, then all or none have degree seven. The proof has two parts. First, a realization of minimal degree seven has defect at most four, the defect being the length of the zero scheme of a minimal section. This uses the residual inflection quartic of the associated degree-seven foliation, local inflection identities, and fixed-point theory for critically fixed rational maps of degree seven with exact finite certificates. Second, the uniformity condition holds at these defects. For multiplicities at most six, a critical realization contains a twelve-line core of minimal weight, and an exact search through extensions of the cores shows that the lattice has a single Galois orbit of realizations; the only family left at defect four has no critical member. For a point of multiplicity seven we use the zero scheme of the pencil derivation and a classification of transversal structures. Terao's conjecture for seventeen and eighteen lines needs only defects one, two and three and, for seventeen lines, this classification at multiplicity seven. The proof is computer-assisted.

math.AG↗

The generic splitting type of the logarithmic bundle of an arrangement of at most sixteen lines is combinatorial

Let $\mathcal{A}$ be an arrangement of $d$ lines in $\mathbb{P}^2_\mathbb{C}$ and let $\mathcal{T}\langle\mathcal{A}\rangle$ be its bundle of logarithmic vector fields. On a general line, $\mathcal{T}\langle\mathcal{A}\rangle$ splits as $\mathcal{O}(-k)\oplus\mathcal{O}(k+1-d)$ with $k=\min\{\mathrm{mdr}(\mathcal{A}),\lfloor\frac{d-1}{2}\rfloor\}$, where $\mathrm{mdr}$ is the minimal degree of a Jacobian syzygy. We prove that $k$ is determined by the intersection lattice whenever $d\le16$, answering in this range a question of Cook, Harbourne, Migliore and Nagel. Since freeness can be read off from $k$ and $c_2(\mathcal{T}\langle\mathcal{A}\rangle)$, this strengthens Terao's freeness conjecture in this range; in particular we obtain a proof of Terao's conjecture for $14$ lines that uses it only for $13$ lines. For any number $d$ of lines, it follows that $\mathrm{mdr}$ is determined by the lattice whenever $\mathrm{mdr}\le6$ and $2\mathrm{mdr}\le d-3$, and that Terao's conjecture holds for every free arrangement whose smaller exponent is at most $7$. The proof reduces the question to critical arrangements: $2s+3$ lines, each containing at most $s+1$ intersection points, with $\mathrm{mdr}=s$. For these, the inflection polynomial of the minimal derivation factors as $Q_\mathcal{A} R$ with $R\in H^0(\mathcal{I}_Z^2(s-3))$, where $Z$ is the zero scheme of the derivation, of length $ν=c_2(\mathcal{T}\langle\mathcal{A}\rangle)-s(s+2)$. For $s\le5$ this forces $ν\le1$, and Chern class arguments then reduce everything to Terao's conjecture for $13$ lines. For $s=6$ we show $ν\le3$, using a Riccati equation along multiple components of $R$, Ziegler's multirestriction, and local computations verified by computer algebra; an exhaustive enumeration of the 26,265,318 candidate lattices with $ν\in\{2,3\}$ that satisfy Bojanowski's inequality completes the proof.

math.AG↗

The recent counterexamples to the Eisenbud-Schreyer-Weyman Ulrich problem (2003) and to Hochster's small Cohen-Macaulay conjecture (1973): a survey, with a generalization to arbitrary dimension

In September 2026 four preprints and an AI-generated manuscript moved one numerical obstruction from projective geometry to local algebra. For a smooth complex projective surface $Y$ with an ample, globally generated divisor $H$ such that $H^2<σ(Y)$, the signature, one obtains in turn: no Ulrich bundles when $H$ is very ample (Anghel, arXiv:2609.02718; extended to every dimension by Lopez, arXiv:2609.06678), answering negatively the Eisenbud-Schreyer-Weyman question of 2003 for sheaves supported on the whole variety; no arithmetically Cohen-Macaulay bundles, i.e. section rings without graded maximal Cohen-Macaulay modules in characteristic zero (Anghel, arXiv:2609.15589); and complete local domains without small Cohen-Macaulay modules, counterexamples to Hochster's 1973 conjecture, if $K_Y\equiv4H$ (Chen, arXiv:2609.24142, proposed) or $σ>2H^2$ (OpenAI, released 6 October). We recall the definitions and main statements, and add two things. First, through a finite projection to $\mathbb P^2$ and an identity already present in the OpenAI manuscript, the three obstructions become one variance identity; for surfaces this explains the factor $2$ between the graded bound $σ\le H^2$ and the local $σ\le2H^2$. Second, we generalize the argument of the OpenAI manuscript to every dimension $m\ge2$, with complete proofs: if $X$ is a smooth complex projective variety of dimension $m$ and $H$ an ample, globally generated divisor, a small Cohen-Macaulay module over the completed section ring of $(X,H)$ forces $(K_X^2-2c_2(X))\cdot H^{m-2}\le(m+4)H^m$. Consequently the completed section rings of all Hesse pairs, including the ball quotient of the first preprint, have no small Cohen-Macaulay module, and products of Hesse surfaces with projective spaces give complete local domains with an isolated singularity, of every dimension $\ge3$, without small Cohen-Macaulay modules.

math.AG↗

Terao's freeness conjecture for arrangements of 15 and 16 lines

We prove that Terao's freeness conjecture holds for arrangements of 15 and of 16 lines in the complex projective plane. Together with the theorem of Barakat and Kühne for at most 14 hyperplanes, this shows that freeness of a central arrangement of at most 16 planes in $\mathbb{C}^3$ is determined by its intersection lattice. The proof has a theoretical and a computational part. Combining published combinatorial criteria for freeness (due to Abe, Dimca, Dimca-Kühne-Pokora and Yoshinaga) with Bojanowski's Hirzebruch-type inequality and an elementary counting identity, we show that a counterexample would have exponents $(6,8)$ or $(7,7)$ for 15 lines and $(7,8)$ for 16 lines, and an intersection lattice subject to strong numerical constraints. We then enumerate all 22,544,280 simple rank 3 matroids satisfying these constraints, up to isomorphism, and determine the complex realization space of each of them exactly over $\mathbb{Q}$. Exactly eleven of them are realizable over $\mathbb{C}$, and on each of these the freeness of the realizations is constant. As a by-product we classify the free arrangements of 15 and 16 lines that are not divisionally free, have maximal multiplicity smaller than the smaller nontrivial exponent, and have at least four intersection points on every line.

math.AG↗

Polarized varieties without arithmetically Cohen-Macaulay bundles and section rings without graded maximal Cohen-Macaulay modules in characteristic zero

We exhibit smooth polarized surfaces $(Y,H)$ over $\mathbf{C}$ carrying no nonzero arithmetically Cohen-Macaulay bundle of any rank. Equivalently, their section rings are three-dimensional normal $\mathbb{N}$-graded $\mathbf{C}$-domains, with an isolated singularity at the vertex, admitting no nonzero finitely generated graded maximal Cohen-Macaulay module. To the author's knowledge no such ring was previously known. Their existence contrasts with the theorem of Hartshorne, Hochster and Peskine-Szpiro that a three-dimensional $\mathbb{N}$-graded domain over a perfect field of characteristic $p>0$ always has one. As a consequence, in characteristic zero Hochster's small Cohen-Macaulay conjecture admits no graded refinement. The nonexistence criterion is numerical: for $|H|$ base-point-free with finite morphism, $H^2<K_Y^2-8χ(\mathcal{O}_Y)=τ(Y)$, the signature of $Y$ over $\mathbf{C}$, leaves no nonzero $\mathcal{E}$ with $H^1(Y,\mathcal{E}(tH))=0$ for all $t$. For Ulrich bundles, being semistable, this is Bogomolov's inequality; what is new is that no stability hypothesis and no condition on the Hilbert polynomial are needed, the Harder-Narasimhan filtration of $\mathcal{E}$ being compared instead with the Horrocks splitting of its direct image on $\mathbb{P}^2$. Hirzebruch's Hesse surfaces with $H_n=4A_n-E_n$, $n\ge3$, qualify. In dimension $m$ the obstruction reads $(m+1)H^m\ge(K_X^2-2c_2(X))\cdot H^{m-2}$, the inequality Lopez proved for Ulrich bundles; products and complete intersections inside them give examples in every dimension $\ge3$, and the threefold examples show that the characteristic-$p$ hypothesis in the criterion of Shimomoto-Tavanfar is essential.

math.AG↗

Smooth Counterexamples to the Eisenbud--Schreyer--Weyman Ulrich Existence Problem

In their 2003 article in the Journal of the American Mathematical Society, Eisenbud, Schreyer and Weyman asked whether every embedded projective variety carries an Ulrich sheaf. In 2017 Beauville proposed a numerical route toward a surface with no Ulrich bundles: in Picard rank one, existence forces $H^2\ge K_S^2-8χ(\mathcal O_S)$, suggesting a search near the Bogomolov--Miyaoka--Yau boundary. We show that the Picard-rank-one hypothesis is not needed for the obstruction: a rank-independent Bogomolov--Hodge argument gives the same inequality on every smooth polarized surface. Using additional Neron--Severi directions on Hirzebruch--Kummer resolutions, the exponent-$3$ Hesse surface admits a very ample class $H=4A-E$ with $H^2=7\cdot3^9<16\cdot3^9=K_Y^2-8χ(\mathcal O_Y)$, hence a smooth counterexample to the Eisenbud--Schreyer--Weyman problem in its standard formulation. Consequently its Chow form has no ESW-type linear determinantal representation arising from an Ulrich sheaf on the embedded surface. Moreover, for every $n\ge3$ the Hesse pair $(Y_n,4A_n-E_n)$ is a counterexample, the surfaces are pairwise non-isomorphic, and $H_n^2/σ(Y_n)=7/(3n^2-11)\to0$, while $K_{Y_n}^2/χ(\mathcal O_{Y_n})\to60/7\approx8.5714$. A general arrangement-theoretic Rees-algebra/Segre mechanism yields further infinite families.

math.AG↗

Relative Logarithmic AKSZ Descent on Joyce Generalized Corners

We study resolution-independent relative logarithmic source complexes for AKSZ-BV-BFV theory on face-oriented Joyce manifolds with generalized corners. Under the geometric hypotheses (G1)-(G3), the Dupont-Panzer-Pym total relative logarithmic complexes of admitted smooth monoidal resolutions represent $j_!Ω_{X^\circ}^\bullet$ and carry the same compactly supported derived trace. The comparison is deliberately unfiltered: it identifies total relative cocycle classes through the common interior, but not individual resolved faces, separate BFV descendants, nonlinear mapping spaces, or absolute regularized integrals. For the intrinsic $b$-source ${}^{b}T[1]X$ we formulate the additional Stokes and transfer data needed to recover strict facewise structures. A multiplicative regularized Stokes trace system on Joyce's face category gives the presymplectic intrinsic BV-BFV identity and incidence descent. The finite datum (G4) is a linear subdivision/aggregation transfer criterion; (G5) is a separate cyclic field-level criterion for classical abelian BF theory. Neither is a general existence theorem, and no general strict multiplicative intrinsic trace is constructed here. We prove the interval contractions needed for codimension-two collars and analyze the positive real conifold through the common star refinement of its two diagonal resolutions. On an explicit finite product-Whitney logarithmic class the exceptional square satisfies (G4) after normal-face totalization, and the transferred differential is exactly the signed intrinsic incidence differential. Its finite algebraic BF dual gives a cyclic logarithmic-cellular shadow, but not the full continuum intrinsic $b$-de Rham datum (G5). General nonlinear continuum pushforward, existence of the intrinsic multiplicative trace, and loop-level logarithmic graph integrals remain open.

math.SG↗

A no-go theorem for special Ulrich bundles, with a complement on primary Burniat surfaces

Let $X$ be a smooth projective surface with $p_g=0$, and let $H$ be an ample divisor with $h^0(\mathcal{O}_X(H))\neq0$, $χ(\mathcal{O}_X(H))\ge q$, and $h^1(\mathcal{O}_X(H))\neq0$. We prove that no rank two bundle $\mathcal{E}$ with $c_1(\mathcal{E})=3H+K_X$, with the Ulrich value of $c_2(\mathcal{E})$, and satisfying $h^0(\mathcal{E}(-H))=0$, can arise from an extension $0 \to \mathcal{O}_X(H+K_X) \to \mathcal{E} \to \mathcal{O}_X(2H)\otimes\mathcal{I}_Z \to 0$. Thus, for this natural Cayley-Bacharach construction, non-speciality of the polarization is necessary rather than merely convenient. We then study primary Burniat surfaces. We show that every ample and base point free divisor is non-special; consequently, every polarization carries a stable special Ulrich bundle of rank two, and the surface is strictly Ulrich wild with respect to every polarization. We also locate the special ample classes on three numerical rays through $K_X$, compute explicit families on these rays, and analyze a twisted-kernel variant of the construction. The degree bound underlying the non-speciality result overlaps with recent work of Y. Cho, while the global consequences and the no-go theorem are independent.

math.AG↗

AKSZ Descent on Manifolds with Ordinary Corners

Under an explicit formal mapping-space hypothesis, we develop a facewise formulation of the classical AKSZ construction on compact oriented manifolds with ordinary corners. The codimension-$r$ data---a mapping space carrying a closed two-form of degree $r-1$, an action of degree $r$, and a cohomological vector field---and the modified Batalin--Vilkovisky/Batalin--Fradkin--Vilkovisky Hamiltonian identity relating consecutive strata are those of the maximally extended BV--BFV theory of Cattaneo--Mnev--Reshetikhin. What is added here is the organization over the entire face poset: the Hamiltonian defect on a face is the sum of the pullbacks of the primitives on its codimension-one faces, weighted by the orientation incidence numbers, so that the boundary term of the single-stratum identity is resolved into its connected pieces with signs. Organizing these defects by the face incidence complex yields a total-complex theorem: factorially normalized facewise transgression is a cochain map, so closed target forms transgress to cocycles, and the twice-iterated defect vanishes because the signed face differential squares to zero. We verify all four codimension-two cancellations explicitly for four-dimensional BF theory on $M=Γ\times[0,1]^2$. We also establish a reduction criterion for singular corner data. If a raw codimension-two descendant is presymplectic and its reduced Dirac structure is the graph of a Poisson bivector, the shifted cotangent construction gives a canonical strict degree-two corner theory. The passage from a reduced Poisson bivector to a strict corner theory is already recorded in \cite{CFT2026}; what is isolated here is the hypothesis under which it applies, and its relation to the face-incidence structure. The construction provides a rigorous ordinary-corner benchmark for extensions of AKSZ descent to Joyce generalized corners.

math.SG↗

On the vector bundles from Chang and Ran's proof of the unirationality of $\mathcal{M}_g$, $g \leq 13$

We combine the idea of Chang and Ran [Invent. Math. 76 (1984), 41-54] of using monads of vector bundles on the projective 3-space to prove the unirationality of the moduli spaces of curves of low genus with our classification of globally generated vector bundles with small first Chern class $c_1$ on the projective 3-space to get an alternative argument for the unirationality of the moduli spaces of curves of degree at most 13 (based on the general framework of Chang and Ran).

math.AG↗

Globally generated vector bundles with $c_1 = 5$ on $\mathbb{P}^n$, $n \geq 4$

We complete the classification of globally generated vector bundles with small $c_1$ on projective spaces by treating the case $c_1 = 5$ on $\mathbb{P}^n$, $n \geq 4$ (the case $c_1 \leq 3$ has been considered by Sierra and Ugaglia, while the cases $c_1 = 4$ on any projective space and $c_1 = 5$ on $\mathbb{P}^2$ and $\mathbb{P}^3$ have been studied in two of our previous papers). It turns out that there are very few indecomposable bundles of this kind: besides some obvious examples there are, roughly speaking, only the (first twist of the) rank 5 vector bundle which is the middle term of the monad defining the Horrocks bundle of rank 3 on $\mathbb{P}^5$, and its restriction to $\mathbb{P}^4$. We recall, in an appendix, from our preprint [arXiv:1805.11336], the main results allowing the classification of globally generated vector bundles with $c_1 = 5$ on $\mathbb{P}^3$. Since there are many such bundles, a large part of the main body of the paper is occupied with the proof of the fact that, except for the simplest ones, they do not extend to $\mathbb{P}^4$ as globally generated vector bundles.

math.AG↗

Globally generated vector bundles with $c_1 = 5$ on $\mathbb{P}^3$

We provide a classification of globally generated vector bundles with $c_1 = 5$ on the projective 3-space. The classification is complete (except for one case) but not as detailed as the corresponding classification in the case $c_1 = 4$ from our paper [Memoirs A.M.S., Vol. 253, No. 1209 (2018), also arXiv:1305.3464]. We determine, at least, the pairs of integers $(a , b)$ for which there exist globally generated vector bundles on the projective 3-space with Chern classes $c_1 = 5$, $c_2 = a$, $c_3 = b$ (except for the case $(12 , 0)$ and the complementary case $(13 , 5)$ which remain undecided), we describe the Horrocks monads of these vector bundles and we organize them into several families with irreducible bases. We use some of the results from our paper [arXiv:1502.05553] (for which we give, however, a direct self-contained proof in one of the appendices of the present paper) to reduce the problem to the classification of stable rank 3 vector bundles $F$ with $c_1(F) = -1$, $2 \leq c_2(F) \leq 4$, having the property that $F(2)$ is globally generated. We use, then, the spectrum of such a bundle to get the necessary cohomological information. Some of the constructions appearing in the present paper are used (and reproduced, for the reader's convenience) in another paper of ours [arXiv:1711.06060] in which we provide an alternative to Chang and Ran's proof of the unirationality of the moduli spaces of curves of degree at most 13 from [Invent. Math. 76 (1984), 41--54].

math.AG↗

Geometry of the Sasakura bundle

The Sasakura bundle is a relatively recent appearance in the world of remarkable vector bundles on projective spaces. In fact, it is connected with some surfaces in $\mathbb P^4$ which missed in early classification papers. The aim of this note is to present various, scattered in the literature, aspects concerning the geometry of this bundle. The last part will be devoted to the place of this bundle in the classification of globally generated locally free sheaves with $c_1 \leq 4$ on $\mathbb P^n$ in a joint paper with I. Coanda and N. Manolache.

math.AG↗

A stable version of Terao conjecture

The aim of this note is to introduce a stable version of Terao conjecture, using the notion of infinitely stably extendability of vector bundles on $\mathbb P^n$, considered and characterized by I. Coanda in arXiv:0907.4040.

math.AG↗

Globally Generated Vector Bundles on P^n with c_1=4

We classify globally generated vector bundles on the projective n-space with first Chern class = 4. This extends previous results for first Chern class at most 3, namely for 2 of Sierra and Ugaglia [J. Pure Appl. Algebra 213 (2009), 2141-2146] and for 3 of Anghel and Manolache [arXiv:1202.6261] and, independently, of Sierra and Ugaglia [arXiv:1203.0185]. It turns out that the case first Chern class = 4 is much more involved than the previous cases, especially on the projective 3-space. Among the bundles appearing in our classification one can find the Sasakura rank 3 vector bundle on the projective 4-space (suitably twisted). In the new version Sections 1, 2, 4. 5, 6 and 7 have been rewritten and some arguments and the presentation have been, hopefully, improved.

math.AG↗

Locally Cohen-Macaulay space curves defined by cubic equations and globally generated vector bundles

We classify globally generated vector bundles with first Chern class $c_1$ at least 4 on the projective 3-space with the property that $E(-c_1+3)$ has a non-zero global section. This (seemingly) technical result allows one to reduce the classification of globally generated vector bundles with $c_1$ at most 7 on the projective 3-space to the classification of stable rank-2 reflexive sheaves with the same properties. The proof is based on a description of the monads of all locally Cohen-Macaulay space curves defined by cubic equations. We extend then this kind of classification to higher dimensional projective spaces. We use this extension to recuperate quickly the classification of globally generated vector bundles with $c_1=4$ on the projective $n$-space for $n$ at least 4, which is part of the main result of our previous paper [arxiv:1305.3464]. We provide, in the appendices to the paper, graded free resolutions for the homogeneous ideals and for the graded structural algebras of all non-reduced locally Cohen-Macaulay space curves of degree at most 4.

math.AG↗

Four generated 4-instantons

We show that there exist mathematical 4-instanton bundles F on the projective 3-space such that F(2) is globally generated (by four global sections). This is equivalent to the existence of elliptic space curves of degree 8 defined by quartic equations. There is a (possibly incomplete) intersection theoretic argument for the existence of such curves in D'Almeida [Bull. Soc. Math. France 128 (2000), 577-584] and another argument, using results of Mori [Nagoya Math. J. 96 (1984), 127-132], in Chiodera and Ellia [Rend. Istit. Univ. Trieste 44 (2012), 413-422]. Our argument is quite different. We prove directly the former fact, using the method of Hartshorne and Hirschowitz [Ann. Scient. Ec. Norm. Sup. (4) 15 (1982), 365-390] and the geometry of five lines in the projective 3-space.

math.AG↗