betaFeedback

WorksMathematics

Log abundance in characteristic zero

Abstract

We prove the rational-boundary log abundance conjecture in every dimension over algebraically closed fields of characteristic zero: every nef ℚ-Cartier log canonical divisor on a projective log canonical pair with effective rational boundary is semiample. Over ℂ, the proof also establishes canonical nonvanishing for smooth projective varieties in every dimension.

Introduction

Let (X,B)(X, B) be a projective log canonical pair with rational boundary. Its log canonical divisor KX+BK_X + B records the canonical class together with the boundary. A rational Cartier divisor is nef if its degree on every curve is nonnegative. It is semiample if a positive Cartier multiple is generated by global sections. Such a multiple defines a morphism, so semiampleness turns numerical positivity into a fibration described by sections. The log abundance conjecture asks whether a nef log canonical divisor is semiample. We give a positive answer in characteristic zero.

Theorem 1.1 (Log abundance). Let (X,B)(X, B) be a normal projective log canonical pair over an algebraically closed field of characteristic zero. Assume that BB is an effective rational divisor and KX+BK_X + B is Q\mathbb{Q}-Cartier. If KX+BK_X + B is nef, then some positive Cartier multiple of it is generated by global sections.

The proof proceeds by dimension induction over C\mathbb{C}. Its inductive assertion is stronger than the displayed theorem: every projective lc pair with effective real boundary and pseudo-effective real Cartier adjoint has a good log minimal model, on which the adjoint is semiample. Smooth canonical nonvanishing asks whether a pseudo-effective canonical class on a smooth variety has a nonzero pluricanonical section. We establish this statement in each dimension before that good-model assertion is used in the same dimension. At the end, descent of the evaluation map of an actual Cartier line bundle gives the theorem over every algebraically closed field of characteristic zero.

We write κ(L)\kappa(L) for the Iitaka dimension of a rational divisor LL, with value −∞-\infty if all positive integral systems are empty. The consequences fall into three groups. Good models, rational-linear triviality when κ=0\kappa= 0, and gluing and extension across boundaries are treated in Section 11. Finite generation of the algebras of rounded pluricanonical sections is proved in Section 12, both absolutely in characteristic zero and for proper morphisms over algebraic bases over C\mathbb{C}. Section 13 relates canonical nonvanishing to rational curves, cotangent tensors, fundamental groups, and quasi-projective covers. These sections give the full statements, additional hypotheses, and derivations from the work of Fujino–Gongyo, Boucksom–Demailly–Păun–Peternell, Lazić–Peternell, Brunebarbe–Campana, Campana, Taji, and Claudon–Höring–Kollár.

There is also a relative analytic application of Fujino’s theorem: Section 14 treats projective surjective morphisms of normal complex analytic varieties with rational lc adjoint nef over the whole base. It obtains a relatively generated Cartier multiple near each compact base subset; the multiple may depend on that subset. This is distinct from the compact Kähler symmetric-pluriform criterion in Section 13; the cotangent-invariant and fundamental-group consequences there remain projective.

Historical development

The threefold proofs show why nonvanishing and boundary geometry are central to abundance. Miyaoka proved the numerical-dimension-one case by studying an effective pluricanonical divisor and its infinitesimal neighborhoods after suitable coverings [50]. Kawamata proved the remaining numerical-dimension-two case for minimal threefolds [38], Theorem 3.1. His Section 4 gives an alternative proof of Miyaoka’s case through finite covers, Hodge theory and compatible infinitesimal deformations. Keel, Matsuki and McKernan established the logarithmic theorem [39, 48, 40]. Fujino extended abundance to semi-log-canonical threefolds [21], where sections on the normalization must agree along the conductor before they define sections on the original space. This compatibility is also needed in higher-dimensional induction: the coefficient-one boundary can be reducible even when the ambient variety is normal.

In higher dimensions, Birkar, Cascini, Hacon and McKernan established minimal models for klt pairs in the big-boundary or log-general-type range [6]. Their finite-generation theorem for projective rational klt pairs does not require bigness [6], Corollary 1.1.2. The rational lc finite-generation consequence below goes beyond that klt setting. Two distinctions remain essential on the boundary of the positive cone. First, nonvanishing produces sections but does not by itself assert that a nef adjoint is semiample. Hashizume reduces real-boundary log canonical nonvanishing and ordinary minimal-model existence through dimension nn to smooth canonical nonvanishing in dimension nn [36], Theorem 1.4]. Our induction must then make these models good. For positive Kodaira dimension, the required step is supplied by lower-dimensional good models [28], Lemma 3.5]. Second, Kodaira dimension zero and numerical dimension zero are different hypotheses. Gongyo proves that a numerically trivial rational log canonical adjoint is Q\mathbb{Q}-linearly trivial [30], Theorem 1.2]. In the zero-Iitaka argument below, numerical triviality forces an already available effective representative to vanish.

Recent work of Liu and Xu treats numerical dimension at most one. Their Theorem 5.1 gives good minimal models for projective log canonical pairs of dimension at most five with nonnegative invariant Kodaira dimension and numerical dimension at most one; their Theorem 5.4 gives a reduction in that numerical range to smooth nonvanishing [47]. The induction here has no numerical-dimension restriction. The ordinary-pair specialization of the companion generalized-minimal-model paper [53], Corollary 1.2] concerns existence of a nef model. It is not an input to this good-model induction. After abundance has been proved, we use that ordinary-pair result in the absolute finite-generation argument of Section 12.

The two tasks in the induction

Assuming good models in dimensions below nn, we first prove smooth canonical nonvanishing in dimension nn, and then make the resulting log minimal models good. Both tasks use a criterion for a signed representative on a reduced boundary. The lower-dimensional hypothesis supplies its whole-boundary semi ampleness assumption. We describe this common ingredient first. For a nef divisor LL, write ν(L)=max⁡{j:LjHn−j>0}\nu(L) = \max\{j : L^jH^{n-j} > 0\}, where HH is ample and n=dim⁡Xn = \dim X. The criterion is

L=KX+D nef,L−cD pseudo-effective (c>0),L = K_X + D\ \text{nef}, \qquad L - cD\ \text{pseudo-effective }(c > 0),
L∼Q∑iaiDi,L∣D semiample⟹κ(L)=ν(L).L \sim_{\mathbb{Q}} \sum_i a_iD_i, \qquad L|_D\ \text{semiample} \quad\Longrightarrow\quad\kappa(L) = \nu(L).

Here (X,D)(X,D) is projective Q\mathbb{Q}-factorial dlt, and D=∑iDiD = \sum_i D_i is the decomposition of the reduced boundary into its prime components. The rational coefficients aia_i may be positive, negative, or zero. Semiampleness is required on the entire reduced scheme DD. Normalization and conductor gluing therefore belong to the induction, as in the slc interface of Fujino–Gongyo [26]. For the good-model task, positive Kodaira dimension is handled by lower-dimensional good models; the signed criterion addresses the remaining zero-Iitaka-dimension case in Proposition 2.5.

The signed criterion is proved geometrically. A generated multiple of L∣DL|_D defines a morphism from DD to projective space. Over a general point of the image of a positive-coefficient component, a root construction separates that component from the negative and coefficient-zero boundary. A filtered Hodge-module calculation lifts local parameters on this image, together with a parameter transverse to the lifted boundary, through all infinitesimal neighborhoods. Fixing the lifted base parameters and varying the transverse one deforms a boundary fiber into a projective subvariety disjoint from DD. Varying the fiber gives a dominating family on which LL is trivial. Section 5 explains the final geometric step: these subvarieties bound the dimension of the nef reduction, and vertical descent identifies LL with the pullback of a big divisor on its base.

This formal deformation method has a direct precedent in Miyaoka’s construction; Miyaoka credits Reid with the suggestion to analyze a pluricanonical divisor [50]. Boundary extension provides a second related approach. Demailly, Hacon and Păun use a singular-metric refinement of Ohsawa–Takegoshi extension to extend pluricanonical sections in the purely log terminal setting [18]. Their nef corollary requires an effective representative whose support contains the coefficient-one divisor and lies in the boundary. Chan and Choi subsequently removed the containment of that representative in the boundary [13]. Here the representative may have both signs, and compatibility is required across the whole reducible boundary. We therefore construct compatible formal lifts on the whole reduced boundary. Semiampleness on that boundary is supplied by the normalization theorem of Fujino and Gongyo [26].

The construction combines familiar tools with an additional compatibility argument. Normalized cyclic covers and their eigenspaces follow the formalism of Esnault and Viehweg [20]. Deligne’s logarithmic Hodge theory supplies the underlying framework [16]. On the singular boundary, we use Saito’s projective strictness and graded de Rham vanishing, together with his comparison with the Du Bois complex [56] and [57]. The proof must still identify the obstruction to lifting and show that these ordinary-variety results apply to the covering construction.

Smooth nonvanishing. To prove smooth nonvanishing, suppose instead that a counterexample exists. The zero-boundary case of logarithmic Iitaka subadditivity, stated below as Theorem 1.2, first excludes positive irregularity through the Albanese map. This is its only use in the proof. The geometric preparations rule out two ways it could be covered by lower-dimensional subvarieties: families that carry a positive current pulled back from a smaller space, and covering families of curves with bounded genus and bounded degree after normalizing the nef divisors used in the construction to a fixed small positive volume. The first exclusion uses positive currents and algebraic webs. The second uses an ascending chain condition for the singularity multiplicities of one fixed current, measured by Lelong numbers at valuations of bounded discrepancy. The signed criterion has a second role here: it forces the full reduced support of any signed canonical representative to have big logarithmic adjoint, as proved in Proposition 6.10. This will control the multisections in the two-slot construction below.

These exclusions produce two incompatible estimates for vanishing at a moving point. At a very general point, the first estimate bounds the vanishing order of every nonzero section of a Cartier multiple kPkP of a suitable nef divisor PP by kk times a small normalized constant. To obtain sections with stronger jet separation, we use the projective bundle associated to the sum of the two pullbacks of a line bundle to the product of a smooth model with itself. Its two summands allow the argument to move in either factor.

Failure of the required jet separation would produce a moving base component. Slice estimates and the curve exclusions leave only components that project generically finitely onto both base factors, hence give correspondences between them. Their branch divisors cannot sweep either factor; a fixed branch complement and bounded degree therefore leave only finitely many covers. After fixing the covers, the correspondences give a family of birational maps whose graphs dominate their product. Hanamura’s birational-group theorem then makes a fixed cover birational to an abelian variety [35] (Theorems 2.1–2.2). The ramification and positive-current argument in Section 6 would give a pluricanonical section on the counterexample, again a contradiction.

Once the projective bundle, its divisors, and the required finite collection of jet sections are fixed, reduction modulo a large prime gives an evaluation map on the product of two copies of the Frobenius-twisted model. The map has full generic rank. Its restriction to the ordinary diagonal is controlled by sections on the Frobenius thickening of the diagonal, where the two relative Frobenius maps agree. A filtration of those sections gives a much smaller rank bound there. The determinant is a section of an external product of two line bundles. The first estimate, expressed by a fixed movable curve on a blowup, bounds vanishing in each factor and hence the order of the determinant at the selected diagonal point. The rank deficit gives a larger lower bound on that order, a contradiction.

Theorem 9.1 states this incompatibility for a fixed smooth variety, fixed jet sections, and a fixed movable-curve bound. Its proof opens Section 9 and uses neither abundance nor subadditivity. The geometric application follows in the same section, using the fixed inputs supplied by the preceding curve exclusions and jet constructions. The numerical subadjunction and moving-kernel results needed for those constructions are proved in Section 7.

The numerical part combines two established methods. The argument that follows a moving base component uses differentiation in the parameter space, as in Ein–Küchle–Lazarsfeld [19] (Proposition 2.3); we give the tracking and adjunction estimates needed here. After reduction to positive characteristic, we compare ordinary jets with Frobenius jets using the local containment recorded by Mustață–Schwede [52] (eq:2.8). The resulting evaluation map is studied through Sun’s canonical Frobenius filtration [59] (Theorem 3.7), whose curve case is treated by Joshi–Raman–Xia–Yu [37] (Section 5.3). We use its description by powers of the diagonal ideal due to Kitadai–Sumihiro [41] (Definition 3.1 and Remark 3.2). Langer’s instability estimate [45] (Corollary 2.5) then controls the slopes of the graded bundles.

The subadditivity input

The all-dimensional argument uses the following completed companion result. Its application occurs only in Lemma 6.1.

Theorem 1.2 (Logarithmic Iitaka subadditivity, [54]). Let f ⁣:X→Yf \colon X \to Y be a surjective morphism with connected fibers between smooth connected projective complex varieties. Let DX,DYD_X,D_Y be reduced effective simple normal crossing divisors, allowing zero boundaries, such that

Supp⁡(f∗DY)⊆Supp⁡(DX).\operatorname{Supp}(f^*D_Y) \subseteq\operatorname{Supp}(D_X).

For a very general smooth fiber FF, put DF=DX∣FD_F=D_X|_F. Then

κ(X,KX+DX)≥κ(F,KF+DF)+κ(Y,KY+DY).\kappa(X,K_X+D_X) \ge\kappa(F,K_F+D_F)+\kappa(Y,K_Y+D_Y).

Here DFD_F is reduced with simple normal crossings and KF+DF∼(KX+DX)∣FK_F+D_F \sim(K_X+D_X)|_F. The convention (−∞)+b=−∞(-\infty)+b=-\infty applies also when b=−∞b=-\infty, and the zero divisor on a point has Iitaka dimension zero.

The support inclusion allows additional components in DXD_X, and the morphism need not be smooth away from the boundary. The theorem has no nefness, abundance, bigness, or good-model hypothesis. Its zero-boundary case is used in the Albanese reduction of Section 6; this is the sole subadditivity input in our proof. The separate characteristic-zero specialization [54] uses a geometric generic fiber. We keep that field-and-fiber statement distinct from the complex, very-general-fiber statement above. No fractional-boundary, orbifold, nonprojective, Hodge-conjectural, or arithmetic extension is assumed here.

The particular zero-boundary application in Lemma 6.1 also follows from the external maximal-Albanese-dimension theorem of Hacon, Popa and Schnell [34]: the smooth base there maps generically finitely to a subvariety of an abelian variety. Cao and Păun prove the underlying abelian-base case [12]. These results cover that application, not the full logarithmic statement above.

Reading order. Section 2 sets out the dimension induction. Sections 3–5 prove the signed-boundary criterion: first the root and adjunction construction, then infinitesimal lifting, and finally compact fibers and descent. Section 6 derives the geometric exclusions for a hypothetical smooth counterexample. After the local tools of Section 7, Section 8 constructs its scalar and two-slot jets. Section 9 proves the independent Frobenius comparison and then constructs its fixed witnesses to establish smooth nonvanishing. Sections 10 and 11 finish the induction, descend actual global generation, and record good models, linear triviality, semi-log-canonical abundance, and supported dlt extension. Section 12 then proves finite generation of rational log canonical rings in the absolute and complex proper relative settings. Section 13 combines canonical nonvanishing and abundance with classical results on rational curves, cotangent tensors, fundamental groups, and quasi-projective covers. Section 14 records relative analytic abundance near compact subsets of the base for projective morphisms. A reader using only the finite-data comparison can start with Theorem 9.1 and its proof in Section 9; neither assumes the counterexample geometry.

Conventions

A variety is integral. Canonical divisors are chosen compatibly on birational models. For a divisor EE on a smooth model p ⁣:W→Xp \colon W \to X, our log discrepancy is

a(E;X,B)=1+ord⁡E(KW−p∗(KX+B)).a(E;X,B)=1+\operatorname{ord}_E(K_W-p^*(K_X+B)).

Log canonicity means nonnegative log discrepancies, and klt means strictly positive log discrepancies. We use the usual dlt and slc conventions. For non-nef pseudo-effective divisors, statements about numerical dimension use Nakayama’s κσ\kappa_\sigma, as specified in the relevant reduction. These notions are not interchanged without a hypothesis that justifies doing so.

For nonempty systems, the Iitaka dimension is the maximum dimension of their images. Numerical and rational-linear equivalence are denoted by ≡\equiv and ∼Q\sim_{\mathbb{Q}}, respectively. Unless a different base field is specified, the constructions are over C\mathbb{C}. A very general point avoids a countable union of proper closed subvarieties. The transfer to arbitrary algebraically closed characteristic-zero fields is made only after the complex argument has been completed.

The inductive framework

The induction has two outputs in each dimension: smooth canonical nonvanishing and good minimal models for real-boundary lc pairs. This section proves the passage from the first output to the second, assuming good models in smaller dimensions. All varieties here are projective over C\mathbb{C}.

For an effective real boundary Δ\Delta, a good log minimal model of (X,Δ)(X,\Delta) is a log minimal model on which the adjoint divisor is semiample. For real divisors, semiampleness means real linear equivalence to the pullback of an ample real divisor by a contraction. We allow the usual definition of a log minimal model in which extracted prime divisors are included in its boundary with coefficient one. For a pseudo-effective divisor AA, we use κσ(X,A)\kappa_\sigma(X,A) for Nakayama’s numerical dimension. When AA is nef this agrees with the intersection-theoretic numerical dimension ν(X,A)\nu(X,A).

Fix an integer n≥1n \ge1.

Assumption 2.1 (Lower-dimensional good models). Every projective log canonical pair of dimension less than nn, with real boundary and pseudo-effective adjoint divisor, has a good log minimal model.

The base case is a point. At the next step, smooth nonvanishing in dimension nn will be proved from Assumption 2.1 in Theorem 9.6. It will then become an input to Proposition 2.5. The boundary-restriction argument below and the special-termination argument in Section 6 are available before this same-dimensional nonvanishing step; they use only the stated lower-dimensional hypothesis.

For clarity, let GLM≤d\mathrm{GLM}_{\le d} denote good-model existence for all projective complex lc pairs with effective real boundary and pseudo-effective adjoint in dimensions at most dd. Let NVn\mathrm{NV}_n denote smooth canonical nonvanishing in dimension nn, and let SAn\mathrm{SA}_n denote semiampleness of nef rational lc adjoints in that dimension. The proof proceeds in the following order; the real-boundary conclusion in the third row supplies the next dimension’s hypothesis.

StageInputOutput and location
BaseDimension zeroGLM≤0\mathrm{GLM}_{\le0}: a point
Smooth stepGLM≤n−1\mathrm{GLM}_{\le n-1} and Theorem 1.2NVn\mathrm{NV}_n, Theorem 9.6
Good-model stepGLM≤n−1\mathrm{GLM}_{\le n-1} and NVn\mathrm{NV}_nGLM≤n\mathrm{GLM}_{\le n}, Proposition 2.5
Nef conclusionGLM≤n\mathrm{GLM}_{\le n}SAn\mathrm{SA}_n, Lemma 2.2

Table 1.

The signed-representative theorem used in both steps is proved in Sections 3–5; its whole-boundary semiampleness hypothesis comes from the lower-dimensional row, via Proposition 2.3.

Comparison and boundary restrictions

We begin with two forms of descent. The first compares the adjoint on a pair and its minimal model as actual divisors on a common resolution. The second joins the sections obtained by adjunction on the components of a reduced dlt boundary.

Lemma 2.2 (Comparison with a nef model). Let (X,Δ)(X,\Delta) be log canonical and let (X′,Δ′)(X',\Delta') be a log minimal model. On a common resolution

X←uW→vX′X \xleftarrow{u} W \xrightarrow{v} X'

there is an effective vv-exceptional divisor FF such that

u∗(KX+Δ)=v∗(KX′+Δ′)+F.(1)u^*(K_X+\Delta)=v^*(K_{X'}+\Delta')+F. \tag*{(1)}

If KX+ΔK_X+\Delta is nef, then F=0F=0. Consequently a nef rational adjoint is semiample whenever it has a good log minimal model. If (X,Δ)(X,\Delta) is klt, its log minimal model is klt and extracts no divisors.

Proof. Put F=u∗(KX+Δ)−v∗(KX′+Δ′)F=u^*(K_X+\Delta)-v^*(K_{X'}+\Delta'). Discrepancy improvement on divisors of XX gives u∗F≥0u_*F\geq0, while −F-F is uu-nef because KX′+Δ′K_{X'}+\Delta' is nef. The negativity lemma therefore gives F≥0F\geq0. At a prime of WW that is not vv-exceptional, the coefficient of FF is zero unless that prime is extracted on X′X'. In the latter case it equals the negative of its log discrepancy over (X,Δ)(X,\Delta). It is therefore nonpositive, and hence zero. Thus FF is vv-exceptional. If the source adjoint is nef, then FF is also vv-nef, so the negativity lemma gives F=0F=0.

The equality of pullbacks transfers semiampleness of a rational adjoint. Indeed a proper birational morphism onto a normal variety has direct image of its structure sheaf equal to that structure sheaf. Projection formula therefore identifies the global sections of each Cartier multiple with the sections of its pullback. If every such pulled-back section vanished over a point downstairs, it could not generate upstairs over that point. Generation upstairs thus implies generation downstairs. Finally, for a klt source the coefficient at an extracted prime would be strictly negative, which is impossible. The remaining discrepancy inequalities show that the model is klt.

Proposition 2.3 (Semi ampleness on the reduced boundary). Assume Assumption 2.1. Let (V,D)(V,D) be a projective Q\mathbb{Q}-factorial dlt nn-fold with DD reduced and M=KV+DM=K_V+D nef. Then M∣DM|_D is semiample as a rational line bundle on the reduced scheme DD.

Proof. The ambient variety VV is klt. The reduced floor of a Q\mathbb{Q}-factorial dlt pair is S2S_2, and it has ordinary double crossings in codimension one; its irreducible components are normal. For the S2S_2 assertion one can work locally with the cyclic class cover of the Q\mathbb{Q}-Cartier divisor DD. This cover is quasi-étale and klt, hence Cohen–Macaulay. The eigensheaf OV(−D)\mathcal{O}_V(-D), being a direct summand of its finite direct image, is Cohen–Macaulay. The divisor sequence then shows that DD is Cohen–Macaulay.

Divisorial adjunction on the normalization ∐Di→D\coprod D_i\to D gives lc pairs (Di,Diff⁡Di(D−Di))(D_i,\operatorname{Diff}_{D_i}(D-D_i)). The different includes the conductor with coefficient one. Removing that conductor contribution defines an effective boundary on DD; together with DD it is an slc pair whose adjoint line bundle is M∣DM|_D. The divisible residue identifications hold in codimension one, with even powers eliminating residue signs at double crossings, and extend by S2S_2.

Each normalized adjoint is nef and semiample by Assumption 2.1 and Lemma 2.2. Semiampleness descends from the normalization by the slc gluing theorem [26], Theorem 1.5, equivalently Theorem 4.3. This gives the required statement on the whole reduced scheme, not merely on its individual components.

A uniform trivial-perturbation argument

We will also perturb a nef adjoint into a klt program. A single Cartier index must survive every step: otherwise the smallness required of the perturbation could change along the program. The following estimate and line-bundle descent give that uniformity.

Lemma 2.4. Let (Z,Δ)(Z,\Delta) be a projective Q\mathbb{Q}-factorial dlt rational pair of dimension nn, and suppose L=KZ+ΔL=K_Z+\Delta is nef. Set P=ΔP=\Delta if the pair is klt and P=⌊Δ⌋P=\lfloor\Delta\rfloor otherwise. Fix m>0m>0 such that mLmL is Cartier. For every rational

0<ϵ<14nm+2,0<\epsilon<\frac{1}{4nm+2},

every step of an LMMP for L−ϵPL-\epsilon P is LL-trivial. On all its models, the transform of LL is nef and its multiple by the same integer mm is Cartier.

References

  1. [1]Florin Ambro. The moduli b-divisor of an lc-trivial fibration. Compositio Mathematica, 141(2):385–403, 2005. Theorem 0.2.
  2. [2]Thomas Bauer, Frédéric Campana, Thomas Eckl, Stefan Kebekus, Thomas Peternell, Sławomir Rams, Tomasz Szemberg, and Lorenz Wotzlaw. A reduction map for nef line bundles. In Complex Geometry: Collection of Papers Dedicated to Hans Grauert, pages 27–36. Springer, Berlin, 2002. Theorem 2.1 in arXiv:math/0106147v1.DOI
  3. [3]Caucher Birkar. On existence of log minimal models. Compositio Mathematica, 146(4):919–928, 2010. Lemma 3.6 in arXiv:0706.1792v3.DOI
  4. [4]Caucher Birkar. On existence of log minimal models II. Journal für die reine und angewandte Mathematik, 658:99–113, 2011. Remark 3.1 and Proposition 3.2(3) in arXiv:0907.4170v1.arxiv.org/abs/0907.4170
  5. [5]Caucher Birkar. Existence of log canonical flips and a special LMMP. Publications Mathématiques de l’IHÉS, 115:325–368, 2012. Theorem 1.9, restated as Theorem 4.1.arxiv.org/abs/1104.4981
  6. [6]Caucher Birkar, Paolo Cascini, Christopher D. Hacon, and James McKernan. Existence of minimal models for varieties of log general type. Journal of the American Mathematical Society, 23(2):405–468, 2010.arxiv.org/abs/math/0610203
  7. [7]Jérémy Blanc. Algebraic structures of groups of birational transformations. In Algebraic Groups: Structure and Actions, volume 94 of Proceedings of Symposia in Pure Mathematics, pages 17–30. American Mathematical Society, 2017. Theorems 3.8–3.9.
  8. [8]Sébastien Boucksom, Jean-Pierre Demailly, Mihai Păun, and Thomas Peternell. The pseudo-effective cone of a compact Kähler manifold and varieties of negative Kodaira dimension. Journal of Algebraic Geometry, 22(2):201–248, 2013. Theorem locators refer to arXiv version 1.
  9. [9]Yohan Brunebarbe and Frédéric Campana. Fundamental group and pluridifferentials on compact Kähler manifolds, 2015. Version 2.arxiv.org/abs/1510.07922
  10. [10]Frédéric Campana. Fundamental group and positivity of cotangent bundles of compact Kähler manifolds. Journal of Algebraic Geometry, 4(3):487–502, 1995.
  11. [11]Frédéric Campana and Mihai Păun. Orbifold generic semi-positivity: an application to families of canonically polarized manifolds. Annales de l’Institut Fourier, 65(2):835–861, 2015.
  12. [12]Junyan Cao and Mihai Păun. Kodaira dimension of algebraic fiber spaces over abelian varieties, 2016. Version 3; Theorem 1.1 gives the abelian-base case.arxiv.org/abs/1504.01095
  13. [13]Tsz On Mario Chan and Young-Jun Choi. Extension with log-canonical measures and an improvement to the plt extension of Demailly–Hacon–Păun, 2021. arXiv:1912.08076v5; Theorem 1.6.1.DOI
  14. [14]Benoît Claudon, Andreas Höring, and János Kollár. Algebraic varieties with quasi-projective universal cover. Journal für die reine und angewandte Mathematik, 679:207–221, 2013.arxiv.org/abs/1102.2762
  15. [15]Michel Coste. Real algebraic sets. In Arc spaces and additive invariants in real algebraic and analytic geometry, volume 24 of Panoramas et Synthèses, pages 1–32. Société Mathématique de France, 2007. Theorem 1.10 in the author’s March 23, 2005 notes.
  16. [16]Pierre Deligne. Théorie de Hodge. II. Publications Mathématiques de l’IHÉS, 40:5–57, 1971.DOI
  17. [17]Jean-Pierre Demailly. Analytic Methods in Algebraic Geometry, volume 1 of Surveys of Modern Mathematics. International Press and Higher Education Press, Somerville, MA, and Beijing, 2012. Theorem numbering refers to the author manuscript dated July 2011.
  18. [18]Jean-Pierre Demailly, Christopher D. Hacon, and Mihai Păun. Extension theorems, non-vanishing and the existence of good minimal models. Acta Mathematica, 210(2):203–259, 2013.arxiv.org/abs/1012.0493
  19. [19]Lawrence Ein, Oliver Küchle, and Robert Lazarsfeld. Local positivity of ample line bundles. Journal of Differential Geometry, 42(2):193–219, 1995. Proposition 2.3.arxiv.org/abs/alg-geom/9408003
  20. [20]Hélène Esnault and Eckart Viehweg. Lectures on Vanishing Theorems, volume 20 of DMV Seminar. Birkhäuser, Basel, 1992.DOI
  21. [21]Osamu Fujino. Abundance theorem for semi log canonical threefolds. Duke Mathematical Journal, 102(3):513–532, 2000.DOI
  22. [22]Osamu Fujino. Introduction to the log minimal model program for log canonical pairs. Unpublished manuscript, 2009. Version 6.01, January 8, 2009; Theorem 3.74(3), p. 109.arxiv.org/abs/0907.1506
  23. [23]Osamu Fujino. Corrigendum to “on subadditivity of the logarithmic Kodaira dimension”. Journal of the Mathematical Society of Japan, 72(4):1181–1187, 2020.DOI
  24. [24]Osamu Fujino. Minimal model program for projective morphisms between complex analytic spaces, 2022. Definitions 2.28, 2.32 and 2.48.arxiv.org/abs/2201.11315
  25. [25]Osamu Fujino. On finiteness of relative log pluricanonical representations. Author manuscript, 2025. Version 0.14, May 29, 2025; Theorem 1.10 and its proof on p. 30.arxiv.org/abs/2506.00760
  26. [26]Osamu Fujino and Yoshinori Gongyo. Log pluricanonical representations and the abundance conjecture. Compositio Mathematica, 150(4):593–620, 2014. Theorem 1.5, restated as Theorem 4.3.DOI
  27. [27]Osamu Fujino and Yoshinori Gongyo. On the moduli b-divisors of lc-trivial fibrations. Annales de l’Institut Fourier, 64(4):1721–1735, 2014.arxiv.org/abs/1210.5052
  28. [28]Osamu Fujino and Yoshinori Gongyo. On log canonical rings. In Higher Dimensional Algebraic Geometry in honour of Professor Yujiro Kawamata’s sixtieth birthday, volume 74 of Advanced Studies in Pure Mathematics, pages 159–169. Mathematical Society of Japan, Tokyo, 2017. Conjectures A–C, Theorem 1.1, Corollary 1.4, and Lemma 3.5.arxiv.org/abs/1302.5194
  29. [29]Osamu Fujino and Kenta Hashizume. Existence of log canonical modifications and its applications. European Journal of Mathematics, 9(1), 2023. Paper No. 13, 32 pp.; Theorem 2.10 in the accepted author version.arxiv.org/abs/2103.01417
  30. [30]Yoshinori Gongyo. Abundance theorem for numerically trivial log canonical divisors of semi-log canonical pairs. Journal of Algebraic Geometry, 22(3):549–564, 2013.arxiv.org/abs/1005.2796
  31. [31]Yoshinori Gongyo and Brian Lehmann. Reduction maps and minimal model theory. Compositio Mathematica, 149(2):295–308, 2013. Read with the numerical-dimension erratum on Lehmann’s research page.DOI
  32. [32]Alexander Grothendieck et al. Revêtements étales et groupe fondamental (SGA 1), volume 224 of Lecture Notes in Mathematics. Springer-Verlag, 1971. Exposé X, Theorem 3.1, and Exposé XII, Theorem 5.1.
  33. [33]Christopher D. Hacon, James McKernan, and Chenyang Xu. Boundedness of varieties of log general type. In Algebraic Geometry: Salt Lake City 2015, Part 1, volume 97.1 of Proceedings of Symposia in Pure Mathematics, pages 309–348. American Mathematical Society, 2018. Published Theorem 4.0.1; Theorem 3.0.1 in arXiv version 2.
  34. [34]Christopher D. Hacon, Mihnea Popa, and Christian Schnell. Algebraic fiber spaces over abelian varieties: around a recent theorem by Cao and Păun, 2017. Version 2; Theorem 1.1.arxiv.org/abs/1611.08768
  35. [35]Masaki Hanamura. Structure of birational automorphism groups, I: non-uniruled varieties. Inventiones Mathematicae, 93:383–403, 1988. Theorems 2.1–2.2; statements also reproduced in Blanc, Theorems 3.8–3.9.DOI
  36. [36]Kenta Hashizume. On the non-vanishing conjecture and existence of log minimal models. Publications of the Research Institute for Mathematical Sciences, 54(1):89–104, 2018. Theorem 1.4 in arXiv:1609.00121v4.arxiv.org/abs/1609.00121
  37. [37]Kirti Joshi, S. Ramanan, Eugene Z. Xia, and Jiu-Kang Yu. On vector bundles destabilized by Frobenius pull-back. Compositio Mathematica, 142(3):616–630, 2006. Section 5.3.arxiv.org/abs/math/0208096
  38. [38]Yujiro Kawamata. Abundance theorem for minimal threefolds. Inventiones Mathematicae, 108:229–246, 1992.DOI
  39. [39]Sean Keel, Kenji Matsuki, and James McKernan. Log abundance theorem for threefolds. Duke Mathematical Journal, 75(1):99–119, 1994. Read together with the authors’ 2004 corrections.DOI
  40. [40]Sean Keel, Kenji Matsuki, and James McKernan. Corrections to “log abundance theorem for threefolds”. Duke Mathematical Journal, 122(3):625–630, 2004.DOI
  41. [41]Yukinori Kitadai and Hideyasu Sumihiro. Canonical filtrations and stability of direct images by Frobenius morphisms. Tohoku Mathematical Journal, 60(2):287–301, 2008. Definition 3.1, Remark 3.2, and Corollary 3.5.DOI
  42. [42]János Kollár. Singularities of the Minimal Model Program, volume 200 of Cambridge Tracts in Mathematics. Cambridge University Press, Cambridge, 2013. With the collaboration of Sándor Kovács.DOI
  43. [43]János Kollár and Sándor J. Kovács. Log canonical singularities are Du Bois. Journal of the American Mathematical Society, 23(3):791–813, 2010.
  44. [44]Sándor J. Kovács and Zsolt Patakfalvi. Projectivity of the moduli space of stable log-varieties and subadditivity of log-Kodaira dimension. Journal of the American Mathematical Society, 30(4):959–1021, 2017. Theorem 9.9 in the author version.DOI
  45. [45]Adrian Langer. Semistable sheaves in positive characteristic. Annals of Mathematics, 159(1):251–276, 2004.DOI
  46. [46]Vladimir Lazić and Thomas Peternell. Rationally connected varieties—on a conjecture of Mumford. Science China Mathematics, 60(6):1019–1028, 2017. Theorem locators refer to arXiv version 1.DOI
  47. [47]Jihao Liu and Zheng Xu. Non-vanishing implies numerical dimension one abundance, 2025. arXiv:2505.05250v2, July 30, 2025; Theorems 3.1, 5.1 and 5.4.arxiv.org/abs/2505.05250
  48. [48]Kenji Matsuki. A correction to the paper “log abundance theorem for threefolds”, 2003. arXiv:math/0302360v1.DOI
  49. [49]V. B. Mehta and A. Ramanathan. Semistable sheaves on projective varieties and their restriction to curves. Mathematische Annalen, 258:213–224, 1982. Theorem 6.1 and Remark 6.2.DOI
  50. [50]Yoichi Miyaoka. Abundance conjecture for 3-folds: case ν = 1. Compositio Mathematica, 68(2):203–220, 1988.
  51. [51]Yoichi Miyaoka and Shigefumi Mori. A numerical criterion for uniruledness. Annals of Mathematics, 124(1):65–69, 1986.DOI
  52. [52]Mircea Mustață and Karl Schwede. A Frobenius variant of Seshadri constants. Mathematische Annalen, 358(3-4):861–878, 2014. Section 2 and Proposition 2.12 in arXiv:1203.1081v1.DOI
  53. [53]OpenAI. Minimal models and Mori fibre spaces for generalized log canonical Q-pairs. OpenAI Math Release preprint OAI:Minimal-models-and-Mori-fibre-spaces-for-generalized-log-canonical-Q-pairs-September-24-2026, 2026. Corollary 1.2.
  54. [54]OpenAI. Orbifold and logarithmic Iitaka subadditivity. OpenAI Math Release preprint OAI:Orbifold-and-logarithmic-Iitaka-subadditivity-September-26-2026, 2026. Corollaries 6.2 and 6.3.
  55. [55]Mihai Păun. Siu’s invariance of plurigenera: a one-tower proof. Journal of Differential Geometry, 76(3):485–493, 2007. Theorem 1.DOI
  56. [56]Morihiko Saito. Mixed Hodge modules. Publications of the Research Institute for Mathematical Sciences, 26(2):221–333, 1990.
  57. [57]Morihiko Saito. Mixed Hodge complexes on algebraic varieties. Mathematische Annalen, 316(2):283–331, 2000. Theorem 0.2, Corollary 0.3 and Proposition 5.3 in arXiv:math/9906088v1.DOI
  58. [58]Yum-Tong Siu. Analyticity of sets associated to Lelong numbers and the extension of closed positive currents. Inventiones Mathematicae, 27:53–156, 1974.DOI
  59. [59]Xiaotao Sun. Direct images of bundles under Frobenius morphism. Inventiones Mathematicae, 173(2):427–447, 2008. Theorem 3.7 and Lemma 4.2 in arXiv:math/0611360v2.DOI
  60. [60]Behrouz Taji. Birational positivity in dimension 4. Annales de l’Institut Fourier, 64(1):203–216, 2014. With an appendix by Frédéric Campana.DOI
  61. [61]The Stacks Project Authors. The Stacks project: Coherent sheaves on Proj, I. https://stacks.math.columbia.edu/tag/01YS, 2026. Tag 01YS, Lemma 30.14.1(5); accessed September 27, 2026.
  62. [62]The Stacks Project Authors. The Stacks project: Grothendieck’s existence theorem, I. https://stacks.math.columbia.edu/tag/087V, 2026. Tag 087V, Lemmas 30.24.1 and 30.24.3; accessed September 26, 2026.

Paper details

vlnc:W_c4f9gy4r
Version
v1 · latest
Published
License
LicenseRef-OpenAI-public
Usage

views · 121

downloads · 5

A view counts once per browser per day, a download each time the PDF is requested.

  • Mathematics

unreleased internal OpenAI model

View statement
Looking for related papers…
Contents