REVIEW 3 major objections 4 minor 43 references
The paper claims a bounded Koszul duality for all infinite-dimensional Koszul algebras, yielding a BGG-type description of coherent sheaf derived categories on Koszul projective schemes.
Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →
For a Koszul quotient Λ of a polynomial ring, the derived category of coherent sheaves on Proj(Λ) is claimed to be equivalent to a quotient of derived categories over the Koszul dual Λ!.
T0 review reviewed 2026-08-04 challenge →
load-bearing objection Plausible program, but Theorem 3.2 relies on an unproved identification and Theorem 4.2 uses an unjustified equality; as written the proofs do not support the claims, though the gaps may be repairable. the 3 major comments →
Koszul Duality for Coherent Sheaves
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
Core claim
The paper's central claim is Theorem 3.2: for every infinite-dimensional Koszul algebra Λ, the Koszul duality functors F and G give mutually inverse triangulated equivalences between the bounded derived category of almost colinear modules over the Koszul dual Λ! and the bounded derived category of almost linear modules over Λ. Theorem 4.2 then claims that when Λ is a Koszul quotient of a polynomial algebra and X = Proj(Λ), the bounded derived category of coherent sheaves is the quotient of D^b(Λ!-AclZ) by D^b(Λ!-CopZ) (a Verdier quotient), equivalently the stabilization S(Λ!-AclZ). Under this identification, coherent sheaves correspond to cohomological shifts of colinear Λ!-modules of infini
What carries the argument
The load-bearing machinery is the Koszul duality pair: the Koszul functor K sends a graded module over the Koszul dual Λ! to a linear complex of projectives over Λ, and the coKoszul functor K∨ goes the other way; their derived extensions F and G provide the equivalence. The boundedness is captured by the 'almost linear' and 'almost colinear' conditions: a module is almost linear (respectively almost colinear) if some degree truncation admits a linear projective (respectively colinear injective) resolution. These conditions mark exactly the modules whose Koszul complex has bounded cohomology, which is what lets the unbounded Koszul duality restrict to bounded derived categories.
Load-bearing premise
The proof of Theorem 3.2 relies on the unproved assertion that any module over the Koszul dual whose Koszul complex has bounded cohomology must be almost colinear (i.e., some truncation is colinear); if that fails, the bounded equivalence and the geometric description of coherent sheaves both collapse.
What would settle it
Exhibit a right-bounded graded module over the Koszul dual of an infinite-dimensional Koszul algebra (for instance, over the Koszul dual of the polynomial ring in two variables) whose minimal injective resolution is not colinear in any degree, yet whose Koszul complex has bounded cohomology. Alternatively, find an almost linear graded module over a Koszul quotient of a polynomial algebra that is not finitely generated, which would break the equality used in Theorem 4.2.
If this is right
- Projective schemes defined by Koszul quotients of polynomial rings get a fully algebraic, noncommutative model for their bounded derived category of coherent sheaves.
- The classical BGG correspondence for projective space is recovered when Λ is a polynomial algebra and Λ! is the exterior algebra.
- For quadratic monomial and certain absolutely Koszul algebras, the duality descends to singularity categories, giving D^b(cqgr(Λ!)) ≅ D_sg(Λ-FpZ).
- For Iwanaga–Gorenstein quadratic monomial algebras, D^b(qgr(Λ!)) and D^b(cqgr(Λ!)) are both equivalent to the stable category of Gorenstein-projective modules.
- The result ties commutative geometry to noncommutative projective geometry by realizing coherent sheaf categories as Verdier quotients of noncommutative derived categories.
Where Pith is reading between the lines
- The identification Λ!-GMod^{-,b} = Λ!-AclZ is the hinge that turns unbounded Koszul duality into the bounded theorem; a natural test is whether a right-bounded non-almost-colinear module can have bounded Koszul cohomology, since that would force the theorem to hold for a strictly smaller category.
- The geometric application relies on the standard theorem for commutative noetherian Koszul algebras that every finitely generated module is almost linear; the same reasoning would need a new input to generalize to noncommutative or non-noetherian Koszul algebras, where almost linear and finitely generated can diverge.
- Because Theorem 4.2 expresses D^b(coh(X)) as a stabilization, it opens a route to computing morphism spaces in the derived category of coherent sheaves from the stable category of almost colinear modules over the Koszul dual.
- The author notes the extension should be formulated in a more general framework without quiver presentations; checking whether the bounded duality survives outside that setting would determine how far the geometric description extends, for example to Grassmannians or Schubert varieties.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper claims a bounded derived Koszul duality for infinite-dimensional Koszul algebras: Theorem 3.2 asserts mutually inverse triangulated equivalences D^b(Λ!-AclZ) ↔ D^b(Λ-AlZ) for every infinite-dimensional Koszul algebra Λ. It then derives singular counterparts, specializes to quadratic monomial algebras (Theorem 3.6) and to absolutely Koszul algebras with an almost-linearity condition (Theorem 3.13), and applies the results to obtain a BGG-type description D^b(coh(X)) ≅ D^b(Λ!-AclZ)/D^b(Λ!-CopZ) ≅ S(Λ!-AclZ) for X=Proj(Λ) with Λ a Koszul quotient of a polynomial algebra (Theorem 4.2).
Significance. If the main equivalence were established, it would provide a uniform bounded Koszul duality beyond the finite-dimensional setting and a geometric BGG correspondence for projective schemes defined by Koszul quotients. The intended framework is natural and the paper is clearly organized. However, the central theorem is not actually proved: the key identification of modules whose Koszul complex has bounded cohomology with almost-colinear modules is asserted without proof, and Theorem 4.2 relies on an unjustified absolute co-Koszulity claim. The paper also depends heavily on the author's own unpublished preprints for several corollaries. These issues prevent the significance from being realized as stated.
major comments (3)
- [§3.1, Theorem 3.2] The proof reduces the theorem to the identification Λ!-GMod^{−,b} = Λ!-AclZ (and dually Λ-GMod^{+,b}=Λ-AlZ). The only cited support, Proposition 3.1, proves one direction: if M is coKoszul then K(M) is exact outside degree 0. It does not prove that H^r(K(M))=0 for r>N implies M_{\le N} is colinear, nor that bounded cohomology of K(M) forces some truncation to be colinear. This converse is exactly the bridge from the unbounded Koszul duality (Theorem 2.1) to the bounded statement. The sentence 'By Proposition 3.1, this is equivalent...' is therefore unsupported. Since Theorem 3.2 is the basis for Corollaries 3.3, 3.4, Theorem 3.13, and Theorem 4.2, this is a load-bearing gap.
- [§4.1, Theorem 4.2] The proof asserts that 'The absolute co-Koszulity of Λ! follows from Lemma 4.1 and Proposition 3.12.' But Proposition 3.12 requires Λ to satisfy both condition (i), absolute Koszulity, and condition (ii), almost-linearity of finitely presented modules. Lemma 4.1 supplies only condition (ii) (Avramov–Eisenbud). Absolute Koszulity of a Koszul quotient of a polynomial algebra is not established and is not a consequence of condition (ii). Thus the claim that Λ! is absolutely co-Koszul for every such Λ is unjustified; it is not known to hold in this generality and is not proved here. This also affects the later identification of objects in the Verdier quotient with weakly co-Koszul modules.
- [§3.1–3.2, Corollaries 3.4 and 3.9] Corollary 3.4 is dismissed with 'The proof is identical to the proof of [9, Theorem 3.16]', and Corollary 3.9 is stated to follow from [9, Theorem 3.17]. The present setting is infinite-dimensional Koszul algebras, whereas [9] treats the finite-dimensional case. No argument is given that the cited proofs carry over verbatim. Combined with the fact that [9]–[13] are the author's own preprints, the paper's main technical content is not self-contained. At minimum, the relevant arguments should be reproduced or the differences addressed.
minor comments (4)
- [§2.2] The notation 'coKoszul' and 'co-Koszul' is used inconsistently; please standardize.
- [Proposition 3.1] The first displayed isomorphism for H^n(K(M)) is stated as 'well known' without a reference; given its importance, a citation or derivation is needed.
- [Theorem 4.2] The phrase 'the category coh(X) identifies with the category consisting of cohomological shifts of colinear Λ!-modules' is imprecise: coh(X) is abelian, not a category of shifts in a triangulated category; clarify the intended embedding.
- [Introduction] The claim D^b(qgr(Λ)) ≃ D^b(coh(X)) by Serre's theorem requires the standing finiteness assumptions, which are not stated at that point in the introduction.
Circularity Check
Central bounded Koszul duality is assembled from an unproved identification asserted via Proposition 3.1; main specializations delegate proofs to same-author preprints.
specific steps
-
self definitional
[Section 3.1, proof of Theorem 3.2; identification of Λ!-GMod^{−,b} with Λ!-AclZ]
"Let M∈Λ!-GMod^{−,b}. By definition, M is right bounded and the complex K(M) has bounded cohomology. Hence there exists an integer N such that H^r(K(M)) = 0 for all r > N. By Proposition 3.1, this is equivalent to saying that the truncation M^{≤N} is colinear."
Proposition 3.1 proves only the forward direction that a coKoszul module has K(M) exact outside degree 0; it does not prove the bounded-converse criterion 'H^r(K(M))=0 for r>N implies M^{≤N} is colinear.' The proof labels this unproved equivalence as the bridge from the unbounded duality to the bounded one. The equality Λ!-GMod^{−,b}=Λ!-AclZ is therefore not derived; it is asserted, and the bounded equivalence D^b(Λ!-AclZ)↔D^b(Λ-AlZ) reduces to that asserted renaming.
-
self citation load bearing
[Section 3.1, Corollary 3.4 (singular Koszul duality), proof]
"Proof.The proof is identical to the proof of [9, Theorem 3.16]. Indeed, under the graded derived Koszul duality, the bounded derived category of finite-dimensional graded Λ!-modules is sent to the full subcategory of perfect complexes in D^b(Λ-GMod^{+,b})."
[9] is Bouhada's own preprint (arXiv:2604.16805) and the paper supplies no proof here; the singular Koszul duality of Corollary 3.4 is asserted to follow by copying that self-cited argument. This is a load-bearing step for the paper's claimed singular duality: without independent verification of [9, Theorem 3.16], the corollary is not derived in this paper.
-
self citation load bearing
[Section 3.2, Lemma 3.5, proof]
"By [11, Lemme 3.2], the second syzygy of M decomposes as a finite direct sum Ω^2M ∼= ⊕_{j=1}^r Λα_j⟨i_j⟩ for suitable arrows α_j and integers i_j. Moreover, by [11, Proposition 3.2], each module Λα_j is linear."
[11] is also by the same author and is cited for the two decisive facts (syzygy decomposition into line modules and linearity of those modules) that make every finitely presented module over a quadratic monomial algebra almost linear. This is the key input for Theorem 3.6, so the monomial specialization of the bounded duality is imported from the author's own preprint rather than proved here.
-
self citation load bearing
[Section 2.4, Theorem 2.1]
"The following theorem is the fundamental result of Koszul duality. It was established in greater generality in [12, 13]; in our setting, it takes the following form. Theorem 2.1 ([12, Theorem 5.16], [13, Theorem 5.7])."
Both [12] and [13] are preprints by the same author (with coauthors), and the proof of Theorem 3.2 starts with 'By Theorem 2.1'. The unbounded equivalence that serves as the base of the whole derivation chain is therefore accepted on the authority of same-author work; no independent proof or machine-check is offered in this paper for this load-bearing input.
full rationale
The paper is not empirically circular: there is no fitted parameter, and the geometric and monomial applications have independent mathematical content once the general bounded duality is granted. However, as written the central Theorem 3.2 is not derived. Its proof asserts the identification Λ!-GMod^{−,b}=Λ!-AclZ through an 'equivalent' that Proposition 3.1 does not establish, and the surrounding technical infrastructure is repeatedly cited to the same author's preprints ([9], [11], [12], [13]). Corollary 3.4's proof is explicitly 'identical to' [9], and Lemma 3.5 is delegated to [11]. Thus the bounded Koszul duality is, in effect, an assumed renaming plus an inheritance from a same-author citation chain, rather than a first-principles derivation. Score 6 reflects partial circularity: the central claim reduces to these inputs by construction, while the specialized applications still contain independent content.
Axiom & Free-Parameter Ledger
axioms (6)
- domain assumption Λ is a Koszul algebra presented by a finite quiver with homogeneous ideal I (not necessarily admissible).
- ad hoc to paper The Koszul duality functors F, G of [12, Theorem 5.16] and [13, Theorem 5.7] are quasi-inverse equivalences D↓(Λ!-GMod−) ↔ D↑(Λ-GMod+).
- ad hoc to paper The results of [9] (e.g., Theorem 2.18, Proposition 2.14, Theorem 3.8) are correct and apply in the present setting.
- ad hoc to paper For quadratic monomial algebras, [11, Lemme 3.2 and Prop 3.2] hold: the second syzygy decomposes into shifts of linear modules Λα_j.
- ad hoc to paper Lemma 3.10: the linear/colinear strands are described by applying the Koszul functor to cohomology; argued to extend [18] to the present setting.
- standard math Lemma 4.1 (Avramov–Eisenbud): every finitely generated graded module over a commutative noetherian Koszul algebra is almost linear.
Cite this review
Pith. "Pith review of Koszul Duality for Coherent Sheaves." pith.science (2026). https://pith.science/paper/IYXH6OX6
@misc{pith2026260714299,
author = {Pith},
title = {Pith review of: Koszul Duality for Coherent Sheaves},
year = {2026},
howpublished = {\url{https://pith.science/paper/IYXH6OX6}},
note = {Machine review of arXiv:2607.14299}
}
read the original abstract
We establish a bounded derived Koszul duality for infinite-dimensional Koszul algebras, and we obtain the corresponding singular Koszul duality. We then apply this framework to two classes of Koszul algebras, namely quadratic monomial algebras and absolutely Koszul algebras satisfying an additional homological condition. For these classes, the general duality specializes to particularly well-behaved forms. As an application to algebraic geometry, let \(\Lambda\) be a commutative noetherian Koszul algebra generated in degree \(1\), and let \(X=\operatorname{Proj}(\Lambda)\). We obtain a Koszul-dual description of the bounded derived category \(\mathsf{D}^{b}\!\bigl(\operatorname{coh}(X)\bigr)\). This gives a BGG-type correspondence for projective schemes defined by commutative noetherian Koszul algebras.
Reference graph
Works this paper leans on
-
[1]
Artin and J
M. Artin and J. J. Zhang,Noncommutative projective schemes, Adv. Math.109(1994), 228–287
1994
-
[2]
L. L. Avramov and D. Eisenbud,Regularity of modules over a Koszul algebra, J. Algebra 153(1992), 85–90
1992
-
[3]
A. A. Beilinson,Coherent sheaves on (P n) and problems of linear algebra, Funct. Anal. Appl.12(1978), 214–216
1978
-
[4]
A. A. Beilinson, V. Ginzburg, and V. V. Schechtman,Koszul duality, J. Geom. Phys.5 (1988), 317–350
1988
-
[5]
Beilinson, V
A. Beilinson, V. Ginzburg, and W. Soergel,Koszul duality patterns in representation theory, J. Amer. Math. Soc.9(1996), no. 2, 473–527
1996
-
[6]
Beligiannis,The homological theory of contravariantly finite subcategories: Auslander– Buchweitz contexts, Gorenstein categories and (co-)stabilization, Comm
A. Beligiannis,The homological theory of contravariantly finite subcategories: Auslander– Buchweitz contexts, Gorenstein categories and (co-)stabilization, Comm. Algebra28 (2000), no. 10, 4547–4596
2000
-
[7]
I. N. Bernstein, I. M. Gel’fand, and S. I. Gel’fand,Algebraic bundles over (P n) and problems of linear algebra, Funct. Anal. Appl.12(1978), 212–214
1978
-
[8]
Bondal and M
A. Bondal and M. Van den Bergh,Generators and representability of functors in commu- tative and noncommutative geometry, Mosc. Math. J.3(2003), no. 1, 1–36
2003
-
[9]
A. M. Bouhada,A non-graded Koszul duality and its applications, preprint (2026), available athttps://arxiv.org/abs/2604.16805
Pith/arXiv arXiv 2026
-
[10]
A. M. Bouhada,Koszul duality for certain finite-dimensional Koszul algebras, in prepara- tion
-
[11]
A. M. Bouhada,Koszul duality for quadratic monomial algebras, preprint (2026), arXiv:2604.20177, available athttps://arxiv.org/abs/2604.20177
Pith/arXiv arXiv 2026
-
[12]
A. M. Bouhada, M. Huang, and S. Liu,Koszul duality for non-graded derived categories, preprint (2019), available athttps://arxiv.org/abs/1908.06153
Pith/arXiv arXiv 2019
-
[13]
A. M. Bouhada, M. Huang, Z. Lin, and S. Liu,A representation theoretic perspective of Koszul theory, preprint (2024), available athttps://arxiv.org/abs/2411.15449
Pith/arXiv arXiv 2024
-
[14]
R. O. Buchweitz,Maximal Cohen–Macaulay Modules and Tate Cohomology over Goren- stein Rings, with appendices by L. L. Avramov, B. Briggs, S. B. Iyengar, and J. C. Letz, Mathematical Surveys and Monographs, vol. 262, American Mathematical Society, Provi- dence, RI, 2021
2021
-
[15]
Chen,Relative singularity categories and Gorenstein-projective modules, Math
X.-W. Chen,Relative singularity categories and Gorenstein-projective modules, Math. Nachr.284(2011), no. 2–3, 199–212
2011
-
[16]
Conca, E
A. Conca, E. De Negri, and M. E. Rossi,Koszul algebras and regularity, inCommutative Algebra: Expository Papers Dedicated to David Eisenbud on the Occasion of His 65th Birthday, I. Peeva (ed.), Springer, New York, 2013, pp. 285–315
2013
-
[17]
Conca, S
A. Conca, S. B. Iyengar, H. D. Nguyen, and T. R”omer,Absolutely Koszul algebras and the Backelin–Roos property, Acta Math. Vietnam.40(2015), no. 3, 353–374. 26
2015
-
[18]
Eisenbud, G
D. Eisenbud, G. Fløystad, and F.-O. Schreyer,Sheaf cohomology and free resolutions over exterior algebras, Trans. Amer. Math. Soc.355(2003), 4397–4426
2003
-
[19]
Eisenbud and S
D. Eisenbud and S. Goto,Linear free resolutions and minimal multiplicity, J. Algebra88 (1984), 89–133
1984
-
[20]
Eisenbud and C
D. Eisenbud and C. Huneke (eds.),Free Resolutions in Commutative Algebra and Algebraic Geometry, A K Peters, Wellesley, MA, 1992
1992
-
[21]
Eisenbud and F.-O
D. Eisenbud and F.-O. Schreyer,Resultants and Chow forms via exterior syzygies, J. Amer. Math. Soc.16(2003), no. 3, 537–579
2003
-
[22]
Eisenbud and F.-O
D. Eisenbud and F.-O. Schreyer,Betti numbers of graded modules and cohomology of vector bundles, J. Amer. Math. Soc.22(2009), no. 3, 859–888
2009
-
[23]
S. I. Gel’fand and Yu. I. Manin,Methods of Homological Algebra, Springer-Verlag, Berlin, 1996
1996
-
[24]
S. B. Iyengar and T. R”omer,Linearity defects of modules over commutative rings, J. Algebra322(2009), no. 9, 3212–3237
2009
-
[25]
Jørgensen,Linear free resolutions over non-commutative algebras, Compos
P. Jørgensen,Linear free resolutions over non-commutative algebras, Compos. Math.140 (2004), no. 4, 1053–1058
2004
-
[26]
Jørgensen,A noncommutative BGG correspondence, Pacific J
P. Jørgensen,A noncommutative BGG correspondence, Pacific J. Math.218(2005), no. 2, 357–377
2005
-
[27]
M. M. Kapranov,On the derived category of coherent sheaves on Grassmann manifolds, Math. USSR-Izv.24(1985), no. 1, 183–192
1985
-
[28]
Kashiwara and P
M. Kashiwara and P. Schapira,Categories and Sheaves, Grundlehren der Mathematischen Wissenschaften, vol. 332, Springer-Verlag, Berlin–Heidelberg, 2006
2006
-
[29]
Mart’inez-Villa and M
R. Mart’inez-Villa and M. Saor’in,Koszul equivalences and dualities, Pacific J. Math.214 (2004), no. 2, 359–378
2004
-
[30]
Mart’inez-Villa and D
R. Mart’inez-Villa and D. Zacharia,Approximations with modules having linear resolutions, J. Algebra266(2003), no. 2, 671–697
2003
-
[31]
Mazorchuk, S
V. Mazorchuk, S. Ovsienko, and C. Stroppel,Quadratic duals, Koszul dual functors, and applications, Trans. Amer. Math. Soc.361(2009), no. 3, 1129–1172
2009
-
[32]
Miyachi,Localization of triangulated categories and derived categories, J
J. Miyachi,Localization of triangulated categories and derived categories, J. Algebra141 (1991), no. 2, 463–483
1991
-
[33]
Orlov,Triangulated categories of singularities and D-branes in Landau–Ginzburg models, Trudy Mat
D. Orlov,Triangulated categories of singularities and D-branes in Landau–Ginzburg models, Trudy Mat. Inst. Steklov.246(2004), 240–262
2004
-
[34]
Orlov,Triangulated categories of singularities, and equivalences between Landau– Ginzburg models, Mat
D. Orlov,Triangulated categories of singularities, and equivalences between Landau– Ginzburg models, Mat. Sb.197(2006), no. 12, 117–132
2006
-
[35]
Orlov,Derived categories of coherent sheaves and triangulated categories of singularities, inAlgebra, Arithmetic, and Geometry: In Honor of Yu
D. Orlov,Derived categories of coherent sheaves and triangulated categories of singularities, inAlgebra, Arithmetic, and Geometry: In Honor of Yu. I. Manin, Vol. II, Progr. Math., vol. 270, Birkh”auser Boston, Boston, MA, 2009, pp. 503–531
2009
-
[36]
Reiten and M
I. Reiten and M. Van den Bergh,Noetherian hereditary abelian categories satisfying Serre duality, J. Amer. Math. Soc.15(2002), no. 2, 295–366. 27
2002
-
[37]
D. Rogalski,An introduction to noncommutative projective geometry, preprint (2014), avail- able athttps://arxiv.org/abs/1403.3065
Pith/arXiv arXiv 2014
-
[38]
Serre,Faisceaux alg’ebriques coh’erents, Ann
J.-P. Serre,Faisceaux alg’ebriques coh’erents, Ann. of Math. (2)61(1955), no. 2, 197–278
1955
-
[39]
Serre,Fiber spaces, in Russian, IL, Moscow, 1957, pp
J.-P. Serre,Fiber spaces, in Russian, IL, Moscow, 1957, pp. 372–453
1957
-
[40]
Shipman,A geometric approach to Orlov’s theorem, Compos
I. Shipman,A geometric approach to Orlov’s theorem, Compos. Math.148(2012), no. 5, 1365–1389
2012
-
[41]
C. A. Weibel,An Introduction to Homological Algebra, Cambridge Studies in Advanced Mathematics, vol. 38, Cambridge University Press, Cambridge, 1994
1994
-
[42]
Yanagawa,Castelnuovo–Mumford regularity for complexes and weakly Koszul modules, J
K. Yanagawa,Castelnuovo–Mumford regularity for complexes and weakly Koszul modules, J. Pure Appl. Algebra213(2009), no. 3, 399–413
2009
-
[43]
Yanagawa,Linearity defect and regularity over a Koszul algebra, Math
K. Yanagawa,Linearity defect and regularity over a Koszul algebra, Math. Scand.104 (2009), no. 2, 273–290. 28
2009
This paper was first reviewed by deepseek-v4-flash on August 4, 2026.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.