Pre-Calabi-Yau algebras and oriented gravity properad

Sergei Merkulov

REVIEW 4 major objections 4 minor 22 references

Pre-Calabi-Yau algebras and oriented gravity properad

T0 review · 4 major / 4 minor · reviewed 2026-08-10 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The oriented gravity properad is a new combinatorial model for the cohomology of moduli spaces of algebraic curves.

desk verdict A plausible and genuinely new oriented ribbon graph properad, but the proof of the main cohomology theorem is carried by omitted computations and an unpublished preprint; worth refereeing, not yet citable as established. read the letter →

arxiv 2501.11158 v2 pith:NCM3DTVB submitted 2025-01-19 math.QA

Sergei Merkulov This is my paper · ORCID

classification math.QA MSC 16E4014H10

keywords pre-Calabi-YaualgebrasorientedribbongraphsproperadsmodulispacesofalgebraiccurvescyclicHochschildcohomologygravityoperadA-infinitygraphcomplexes

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper proves that a differential graded properad built from oriented ribbon graphs, called the oriented gravity properad, has cohomology equal to that of the ordinary ribbon graph properad, and therefore computes the compactly supported cohomology of moduli spaces of algebraic curves with marked points. It also proves that any pre-Calabi-Yau extension of an A-infinity algebra structure makes the dual cyclic Hochschild complex into a module over this properad, and that the minimal resolution of the Lie operad acts on it. The result matters because it connects noncommutative deformation theory to the geometry of curve moduli and gives a purely combinatorial, graph-theoretic replacement for those moduli cohomology spaces.

What carries the argument

The central object is the dg properad ORGraphs^d (and its unshifted predecessor RGraphs^d): a properad of connected ribbon graphs with directed edges, labelled boundaries, labelled white vertices, and unlabelled black vertices, with vertices carrying cohomological degree d and edges degree 1−d. The differential is the standard splitting-of-vertices operation, with certain target vertices killed by passing to an ideal. The argument that carries the main theorem is the interpolating properad ORGraphs_{d,d+1}, which allows both unlabelled white and black vertices; Proposition 4.3 shows the quotient killing unlabelled white vertices is quasi-isomorphic to $ORGraphs^{{d+1}}$, and Proposition 4.4 shows the quotient killing essential black vertices is quasi-isomorphic to RGraphs^d. These two quasi-isomorphisms, proved by filtrations and the acyclicity of elementary two-term complexes, identify the cohomology of the oriented properad with that of the classical ribbon graph complex.

What would settle it

A direct computation of δ•^2 on a small generator of ORGraphs_{d,d+1}, for instance the graph with one labelled boundary, one labelled white vertex, and one unlabelled black vertex, would either confirm or refute the omitted square-zero calculation; if a nonzero value appears, Propositions 4.3 and 4.4 cannot both hold, and the main theorem fails.

Watch

Extended reading notes

Core claim

Main Theorem 1.3 states that for any integers d, m ≥ 1, n ≥ 1, the cohomology of the oriented ribbon graph properad $ORGraphs^{{d+1}}$(m,n) is isomorphic to H•(RGraphs^d(m,n)), which in turn equals the product of compactly supported cohomology groups ∏_{g≥0, 2g+m+n≥3} $H_c^{{•-m+d(2g-2+m+n)}}$(M_{g,m+n}). In words, the new properad of directed ribbon graphs with no closed directed cycles computes the same compactly supported cohomology of moduli spaces M_{g,m+n} as the ordinary ribbon graph complex, up to a degree shift. Theorem 1.1 shows that a degree d pre-Calabi-Yau extension π of an A∞-structure μ on a graded vector space A induces a natural action of ORGraphs^d on the dual cyclic Hochschild complex Cyc•(A,K), and Proposition 3.5.1 supplies a nontrivial morphism from the minimal resolution Holie^d into $ORGraphs^{{d+1}}$, making Cyc•(A,K) a Holie^d-algebra by Corollary 1.2. The paper also proves that the higher Hochschild cohomology of any pre-CY algebra is naturally an algebra over the gravity operad.

Load-bearing premise

The whole result rests on a chain of technical checks that are asserted here and partly deferred to earlier papers (one unpublished): the differential on the interpolating graph complex must square to zero, and certain subcomplexes must have no cohomology.

Editorial extensions

If this is right

The cohomology of ORGraphs^{d+1}(m,n) is known explicitly as the product of compactly supported cohomology groups of moduli spaces M_{g,m+n} with the stated degree shifts.
Every pre-Calabi-Yau extension of an A∞ structure gives a representation of ORGraphs^d on the dual cyclic Hochschild complex, so cyclic cohomology carries these graph operations.
Degree (d+1) pre-Calabi-Yau extensions make the dual cyclic Hochschild complex into a Holie^d-algebra, extending the usual Lie structure to all higher operations.
The higher Hochschild cohomology of any pre-CY algebra is naturally an algebra over the gravity operad, with the binary operation coming from the standard Lie bracket.
The genus-zero part of the cohomology recovers Getzler's gravity operad, so the construction is a full properadic extension of that classical structure.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

Editorial inference: The interpolation argument suggests a general pattern—changing vertex colour and degree in a graph complex can be bridged by a two-colour complex whose quasi-isomorphisms reduce the computation to known complexes; the same scheme may apply to other graph-complex comparisons.
Editorial inference: Because the paper works with coinvariants under cyclic groups, an analogous construction for invariants or for non-cyclic Hochschild complexes might produce oriented properads acting on other natural cohomology theories.
Editorial inference: The explicit verification of the square-zero differential on the interpolating properad is the most direct spot to test the proof; a small computer algebra check on low genus graphs could confirm or refute the asserted acyclicity before relying on the unpublished claims.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Quick take: the paper constructs ORGraphs^d, a directed ribbon graph properad, proves it acts on the dual cyclic Hochschild complex of any pre-CY algebra, and identifies its cohomology with compactly supported cohomology of M_{g,m+n}. If the main theorem holds, this is a genuinely new combinatorial model and a clean cyclic analogue of [KTV]. The action theorem (1.1) is proven in reasonable detail; the Holie morphism is plausible; the gravity-algebra corollary is a nice bonus.

The soft spot is exactly where the reader and stress-test put it: Section 4. The proof of Theorem 1.3 goes through an interpolating properad ORGraphs_{d,d+1} with differential δ = δ• + D_γ. Two identities are load-bearing: δ•^2 = 0 and δ•γ + γ∘_1γ = 0. Both are dismissed as "straightforward" and the second is deferred to [MWW], an unpublished preprint. Since δ^2 = 0 splits exactly into those identities, the main theorem is conditional on computations the paper does not contain. That is a real incompleteness, not a manufactured one.

I would also flag Proposition 4.3's acyclicity argument. The proof decomposes the associated graded complex into factors C_v and says every graph has a vertex of type (ii) because there are no closed directed paths. That is true for the generators, but after properadic composition the property can fail; the argument needs a justification it does not give. The same style of appeal to [M3] and [MWW] appears in 4.4. None of this makes me think the theorem is false—the strategy is coherent and the cited results are the right tools—but as written the central proof is a series of deferred checks.

On circularity: I disagree with any charge that the argument is circular. The comparison is to independently known cohomology of RGraphs^d, and self-citations are used appropriately. The citation pattern is normal for this subfield.

Who this is for: people working on operadic models of moduli spaces, pre-Calabi-Yau algebras, and graph complexes. A specialist referee can probably check Section 4 in a few days; the gaps are tedious-computation-shaped, not conceptual. I would send it to review, and ask the referee to verify δ•^2 = 0 and Lemma 4.1.1 explicitly. I would not cite Theorem 1.3 as established until that check is public.

Referee Report

4 major / 4 minor

Summary. The paper introduces a dg properad ORGraphs^d of oriented ribbon graphs and studies its actions and cohomology. Theorem 1.1 asserts that any degree-d pre-Calabi-Yau extension of an A-infinity structure induces an action of ORGraphs^d on the dual cyclic Hochschild complex. Proposition 3.5.1 gives a morphism from the minimal resolution Holie^d to ORGraphs^{d+1}, yielding Corollary 1.2. The main theorem (Theorem 1.3) computes H^•(ORGraphs^{d+1}(m,n)) as H^•(RGraphs^d(m,n)), which by [C, M1] equals the compactly supported cohomology of moduli spaces M_{g,m+n}. The proof uses an interpolating properad ORGraphs_{d,d+1} with two differentials and two quotient maps to ORGraphs^{d+1} and RGraphs^d.

Significance. If Theorem 1.3 is correct, the paper provides a new combinatorial model for the totality of compactly supported cohomology groups of moduli spaces of algebraic curves, and connects pre-Calabi-Yau structures to cyclic cohomology and the gravity operad. The constructions are explicit and the paper is well-organized, with several detailed proofs such as Lemma 3.2.1 and Proposition 3.4.1. However, the central proof of Theorem 1.3 rests on several asserted but unverified computations and on citations to an unpublished preprint, so the result is presently conditional. The potential significance is high: it would extend the properadic approach to moduli spaces and offer a new setting for actions on cyclic Hochschild complexes.

major comments (4)

[§4.1, displayed equation after Lemma 4.1.1] The differential δ = δ• + D_γ on ORGraphs_{d,d+1} is the backbone of the proof of Theorem 1.3. The identity δ•^2 = 0 is dismissed as 'a straightforward but tedious calculation', and Lemma 4.1.1, which gives the cross-identity δ•γ + γ ∘_1 γ = 0, is stated with proof deferred to the unpublished preprint [MWW]. These identities are load-bearing: without them, (ORGraphs_{d,d+1}, δ) is not a dg properad, and Propositions 4.3 and 4.4 cannot be quasi-isomorphisms of dg properads. The manuscript should supply these computations or give a complete, publicly accessible proof.
[§4.3, Proof of Proposition 4.3] The acyclicity argument for ker π1 considers the associated graded complex gr ker π1^marked and claims that 'Any generating graph Γ has least one vertex of type (ii)'. This claim is used to conclude that each tensor factor CΓ has an acyclic factor. However, properadic composites of graphs can replace an unlabelled target by a graph with no unlabelled targets, so it is not clear that every graph occurring in the associated graded has a vertex of type (ii). This step needs an independent justification, as it is essential for the quasi-isomorphism ORGraphs_{d,d+1} → ORGraphs^{d+1}.
[§3.5, Proposition 3.5.1] The morphism Holie^d → ORGraphs^{d+1} is the basis for Corollary 1.2, but its proof is omitted entirely, described only as 'a straightforward calculation'. Since the map is given by an explicit sum with signs, verifying that it commutes with the differentials is a non-trivial check. The reader cannot verify this claim from the text, and the result is used in the main narrative. The author should provide the calculation or a detailed reference.
[§4.3 and §4.4] The proofs of Propositions 4.3 and 4.4 defer crucial acyclicity claims to [MWW] (an unpublished preprint) and to [M3]. Specifically, the proof of Proposition 4.3 says 'cf. §4.1 in [MWW]' and Proposition 4.4 says 'cf. Lemma 6.2.3 in [M3]', while Lemma 4.1.1 is deferred to [MWW]. This makes the proof of Theorem 1.3 not self-contained and unverifiable in its present form. The author should either prove these claims in the paper or state precisely which published arguments imply them.

minor comments (4)

[§3.2] The notation is inconsistent: the text uses both RGraphsor_d and RGraphs^or_d for the same object, and similarly ORGraphs^d versus ORGraphsd in different sections. A unified notation would improve readability.
[§3.4] The statement 'It is easy to check that ORGraphsd is a dg sub-properad' would benefit from a short verification, since closure under the differential and compositions is not obvious given the quotient by the ideal of black targets.
[§3.2.1, Proof of Lemma 3.2.1] In the infinite-dimensional case, the action is described only 'along the same lines' with substitution of outputs into inputs. A more explicit description of the operation on generators would help the reader verify the claims.
[General] There are numerous typographical errors and formatting artifacts in the arXiv text, such as 'RG raphsor', missing spaces, and garbled diagrams. These should be corrected in the final version.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: ORGraphs^{d+1} cohomology is compared to the independently known RGraphs^d cohomology; deferred calculations are proof-gaps, not reductions to the target.

full rationale

The derivation is not circular. The new properad ORGraphs^{d+1} is constructed from directed ribbon graphs with black/white unlabelled vertices and an explicit differential (Section 3.4, formula (5)); it is not defined as the cohomology of RGraphs^d or of the moduli spaces M_{g,m+n}. Main Theorem 1.3 is proved by introducing an interpolating properad ORGraphs_{d,d+1} and showing that the two projections to ORGraphs^{d+1} and to RGraphs^d are quasi-isomorphisms (Propositions 4.3 and 4.4). The target H^•(RGraphs^d(m,n)) is independently identified with the compactly supported cohomology of M_{g,m+n} via Costello's construction and earlier work [C, M1]. Thus the claimed equality is a theorem established by comparison with known objects, not an input or a rename. The paper does rely heavily on self-citations for technical ingredients: Section 4.1 defers the verification of δ•^2 = 0 and Lemma 4.1.1 (δ•γ + γ∘_1γ = 0) to [M3, MWW], and Propositions 4.3 and 4.4 defer key acyclicity arguments to [MWW] and [M3], respectively. These are load-bearing dependencies and should be checked by referees, especially since [MWW] is an unpublished preprint, but they are not circular: the cited results are prior computational statements, not the cohomological statement being proved. A paper can be non-self-contained without being circular. The appropriate verdict is therefore no significant circularity, with the caveat that the main theorem's completeness depends on the correctness of those deferred computations.

Assumptions & free parameters 0 free parameters · 5 assumptions · 2 invented entities

The construction relies on foundational results from the author's earlier work and collaborators: the twisting functor, the cohomology of the undirected ribbon graph properad, and the gravity operad identification. These are standard in the field but not proven in this paper. The most fragile inputs are the unproved 'straightforward' computations that define the differential on the interpolating properad.

assumptions (5)

domain assumption The twisting endofunctor tw of [Wi1] produces dg properads and preserves quasi-isomorphisms.
Used to define twRGraphs^d and RGraphs^or_d in §3.1-3.2; the paper relies on the properties of tw without proving them.
domain assumption The cohomology of RGraphs^d(m,n) equals ∏_g H_c^{•-m+d(2g-2+m+n)}(M_{g,m+n}) as established in [C, M1].
This is the target of the main theorem and is used verbatim in Theorem 1.3; the paper cites [M1] for the proof.
domain assumption The identification H^•(RTrees^d) ≅ Grav{d-2} from [Wa] holds.
Used in Lemma 3.3.1 to conclude that higher Hochschild cohomology is a gravity algebra.
ad hoc to paper The omitted computations in the paper are correct, specifically δ•^2 = 0 on ORGraphs_{d,d+1} (§4.1), the quasi-isomorphism claims in Props 4.3 and 4.4, and Prop 3.5.1.
These are asserted as 'straightforward calculations' or deferred to [M3, MWW] without full proof in the text.
standard math The field K has characteristic zero.
Explicitly assumed in §1.4 to ensure Maschke's theorem applies in Prop 4.3 proof.

invented entities (2)

ORGraphs^d: a new dg properad of oriented ribbon graphs. independent evidence
purpose: To act on the dual cyclic Hochschild complex Cyc^•(A,K) of a pre-CY algebra and to model cohomology of moduli spaces M_{g,m+n}.
Its cohomology is proven to equal the compactly supported cohomology of moduli spaces M_{g,m+n}, an independently known invariant, providing a falsifiable check of the construction.
ORGraphs_{d,d+1}: interpolating dg properad with two differentials.
purpose: Technical tool to compare ORGraphs^{d+1} with RGraphs^d via two quotient quasi-isomorphisms.
It is introduced solely to prove Theorem 1.3; no external invariant is associated with it beyond the two quotient results, so it has no independent falsifiable handle.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Pre-Calabi-Yau algebras and oriented gravity properad." pith.science (2026). https://pith.science/paper/NCM3DTVB

@misc{pith2026250111158,
  author       = {Pith},
  title        = {Pith review of: Pre-Calabi-Yau algebras and oriented gravity properad},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NCM3DTVB}},
  note         = {Machine review of arXiv:2501.11158}
}

abstract

We study the dual cyclic Hochschild complex $Cyc^\bullet(A,\mathbb{K})$ of a (possibly, infinite-dimensional) $A_\infty$-algebra $(A,\mu)$ and prove that any pre-Calabi-Yau extension $\pi$ of the given $A_\infty$ structure $\mu$ in $A$ induces on the cyclic cohomology of $(A,\mu)$ a representation of a new dg properad of oriented ribbon graphs. We compute the cohomology of that properad in terms of the compactly supported cohomology groups of moduli spaces $\mathcal{M}_{g,m+n}$ of algebraic curves of genus $g$ with $m+n$ marked points. We also show that the gravity operad acts naturally on the higher Hochschild cohomology of any pre-CY algebra $(A, \pi)$.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

22 extracted references · 14 canonical work pages

[1]

Kevin Costello, The A_ operad and the moduli space of curves , arXiv:

work page
[2]

Phys.\ 147 (2020) 103522

Xiaojun Chen, Farkhod Eshmatov and Leilei Liu, Gravity algebra structure on the negative cyclic homology of Calabi-Yau algebras , J.\ Geom. Phys.\ 147 (2020) 103522

work page 2020
[3]

Kevin Costello, A dual version of the ribbon graph decomposition of moduli space , Geometry & Topology 11 (2007) 1637-1652

work page 2007
[4]

Vladimir Dotsenko, Sergey Shadrin, and Bruno Vallette, Maurer-Cartan methods in deformation theory: the twisting procedure , London Mathematical Society Lecture Note Series (488), Cambridge University Press, 2023

work page 2023
[5]

163 (1994), no

Ezra Getzler, Two-dimensional topological gravity and equivariant cohomology , Comm.\ Math.\ Phys. 163 (1994), no. 3, 473-489

work page 1994
[6]

Natalia Iyudu and Maxim Kontsevich, Pre-Calabi-Yau algebras and noncommutative calculus on higher cyclic Hochschild cohomology , arXiv: 2011.11888 (2020)

arXiv 2020
[7]

Pre-CalabiYau algebras as noncommutative Poisson structures , Journal of Algebra 567 (2021), pp

Natalia Iyudu, Maxim Kontsevich, and Yannis Vlassopoulos. Pre-CalabiYau algebras as noncommutative Poisson structures , Journal of Algebra 567 (2021), pp. 63-90

work page 2021
[8]

In Proceedings of the I

Maxim Kontsevich, Formal (non)commutative symplectic geometry . In Proceedings of the I. M. Gelfand seminar 1990-1992, pages 173-188. Birkhauser, 1993

work page 1990

Show all 22 references

[9]

Maxim Kontsevich and Yan Soibelman, Notes on A_ -algebras, A_ -categories and non-commutative geometry. I. In: Homological Mirror Symmetry, Lecture Notes in Physics, volume 757, Springer, 2008

2008
[10]

To appear in European Journal of Mathematics

Maxim Kontsevich, Alex Takeda, and Yiannis Vlassopoulos Pre-Calabi-Yau algebras and topological quantum field theories , preprint arXiv: 2112.14667 (2021). To appear in European Journal of Mathematics

2021 arXiv
[11]

Johan Leray and Bruno Vallette, Pre-Calabi-Yau algebras and homotopy double Poisson gebras , arXiv: 2203.05062 (2022)

2022
[12]

Sergei Merkulov, Gravity prop and moduli spaces _ g,n , arXiv:2108.10644 (2021)

2021
[13]

Sergei Merkulov, Twisting of properads , J.\ Pure and Appl.\ Algebra , 227 , Issue 10 (2023) 1-47

2023
[14]

Sergei Merkulov, On interrelations between graph complexes , Int.\ Math.\ Res.\ Notices, 2025 , Issue 2 (2025), rnae287, https://doi.org/10.1093/imrn/rnae287

2025 doi
[15]

Sergei Merkulov and Thomas Willwacher, Props of ribbon graphs, involutive Lie bialgebras and moduli spaces of curves , preprint arXiv:1511.07808 (2015) 51pp

2015 arXiv
[16]

Sergei Merkulov, Thomas Willwacher and Vincent Wolff, The oriented graph complex revisited , arXiv:2411.19657 (2024)

2024 arXiv
[17]

Thomas Tradler and Mahmoud Zeinalian, Algebraic string operations , K-Theory, 38 (1) (2007) 59-82

2007
[18]

113 (1987), 299-339

Robert Penner, The decorated Teichm\"uller space of punctured surfaces , Comm.\ Math.\ Phys. 113 (1987), 299-339

1987
[19]

Math., 288 (2016) 240-307

Nathalie Wahl and Craig Westerland, Hochschild homology of structured algebras , Adv. Math., 288 (2016) 240-307

2016
[20]

4, 1403-1464

Benjamin Ward, Maurer-Cartan elements and cyclic operads , J.\ Noncommut.\ Geom.\ 10 (2016), no. 4, 1403-1464

2016
[21]

Thomas Willwacher, M.\ Kontsevich's graph complex and the Grothendieck-Teichmueller Lie algebra , Invent. Math. 200 (2015), 671-760

2015
[22]

Thomas Willwacher, Oriented graph complexes , Comm. Math. Phys. 334 (2015), no. 3, 1649-1666

2015

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.

[1] [1]

Kevin Costello, The A_ operad and the moduli space of curves , arXiv:

work page

[2] [2]

Phys.\ 147 (2020) 103522

Xiaojun Chen, Farkhod Eshmatov and Leilei Liu, Gravity algebra structure on the negative cyclic homology of Calabi-Yau algebras , J.\ Geom. Phys.\ 147 (2020) 103522

work page 2020

[3] [3]

Kevin Costello, A dual version of the ribbon graph decomposition of moduli space , Geometry & Topology 11 (2007) 1637-1652

work page 2007

[4] [4]

Vladimir Dotsenko, Sergey Shadrin, and Bruno Vallette, Maurer-Cartan methods in deformation theory: the twisting procedure , London Mathematical Society Lecture Note Series (488), Cambridge University Press, 2023

work page 2023

[5] [5]

163 (1994), no

Ezra Getzler, Two-dimensional topological gravity and equivariant cohomology , Comm.\ Math.\ Phys. 163 (1994), no. 3, 473-489

work page 1994

[6] [6]

Natalia Iyudu and Maxim Kontsevich, Pre-Calabi-Yau algebras and noncommutative calculus on higher cyclic Hochschild cohomology , arXiv: 2011.11888 (2020)

arXiv 2020

[7] [7]

Pre-CalabiYau algebras as noncommutative Poisson structures , Journal of Algebra 567 (2021), pp

Natalia Iyudu, Maxim Kontsevich, and Yannis Vlassopoulos. Pre-CalabiYau algebras as noncommutative Poisson structures , Journal of Algebra 567 (2021), pp. 63-90

work page 2021

[8] [8]

In Proceedings of the I

Maxim Kontsevich, Formal (non)commutative symplectic geometry . In Proceedings of the I. M. Gelfand seminar 1990-1992, pages 173-188. Birkhauser, 1993

work page 1990