REVIEW 1 major objections 4 minor 1 cited by
Tropicalization of $\psi$ classes
T0 review · 1 major / 4 minor · reviewed 2026-08-11 · deepseek-v4-flash
Pith's one-line read For tropicalizable families with enough affine data near a section, the tropical psi class equals the tropicalization of the algebraic psi class.
desk verdict A conditional positive answer to the CGM22 psi-class question, with a new affine tropicalization functor; the main theorem is solid but needs a completeness hypothesis added to Corollary 5.11. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The load-bearing construction is the affine structure on the cone complex of a toroidal variety. A strict piecewise linear function $\phi$ on the open star of a cone $\sigma$ is declared affine at $\sigma$ exactly when the associated line bundle $\mathcal{O}_{X_\sigma}(\phi)$ restricts trivially to the stratum $V(\sigma)$; at a cell at infinity $\sigma/\tau$ one additionally requires $\phi$ to be constant on $\tau$. This turns tropicalization into a functor from toroidal varieties to tropical spaces, invariant under logarithmic modifications. A piecewise linear function is combinatorially principal when it agrees, on every cone, with some affine function; this is exactly what makes $\mathrm{Aff}_{\mathsf{C}}(-\mathsf{s}_i)$ a tropical line bundle rather than a mere pseudo-torsor. The comparison for $\psi$ classes then runs through the sheaf $\mathrm{Aff}_{\mathsf{C}}(-\mathsf{s}_i)$: its local sections are affine functions on the finite part of the tropical curve that approach the $i$-th section with slope $-1$, and Proposition 5.2 identifies $\mathrm{Trop}(\mathcal{O}_{\mathcal{C}}(\mathsf{s}_i))$ with that pseudo-torsor, which Theorem 5.5 pulls back along the section to give the tropical cotangent bundle.
What would settle it
Recompute the tropical $\psi_1$ class on one of the two genus-one admissible-cover families of Section 6 by evaluating the balancing condition with the affine functions pulled back from $\mathsf{M}^{\mathrm{trop}}_{0,5}$; the paper predicts coefficients $2/3$ and $1$ on the rays of type $\rho_a$ and $\rho_b$, and the pushed-forward class $12\rho_{\mathrm{irr}}$. A mismatch with any of these numbers would refute the comparison, while a match would confirm it in a case where the affine structure is only partially available.
Extended reading notes
Core claim
On the paper's own terms, the central discovery is Theorem 5.5 and Corollary 5.11. Given a tropicalizable family $\pi:\mathcal{C}\to\mathcal{B}$ of $n$-marked stable curves with tropicalization $\Pi:\mathsf{C}\to\mathsf{B}$, and given the piecewise linear function $\varphi_i$ on $\mathsf{C}$ that has slope one along the ray dual to the $i$-th section and slope zero on all other rays, the condition that $\mathrm{Aff}_{\mathsf{C}}(-\mathsf{s}_i)$ is a tropical line bundle—equivalently, that $\varphi_i$ is combinatorially principal—forces the tropicalization of the algebraic cotangent line bundle $\mathcal{L}_i$ to be isomorphic to the tropical cotangent bundle $\mathsf{L}_i^{\mathrm{trop}}=\mathsf{s}_i^*\,\mathrm{Aff}_{\mathsf{C}}(-\mathsf{s}_i)$. Taking first Chern classes yields $\mathrm{Trop}(\psi_i)=\psi_i^{\mathrm{trop}}$ as tropical cycles. In words: tropical geometry sees the algebraic $\psi$ class exactly when the family degenerates enough near the section; on that locus the combinatorial $\psi$ classes are the tropicalizations of algebraic ones.
Load-bearing premise
The whole comparison applies only to tropicalizable families, meaning the tropicalization must faithfully reflect the dual graphs of the algebraic fibers; when monodromy hides reducible fibers this fails and an étale cover is required before the theorem can be used.
Editorial extensions
If this is right
- On the tropicalizable locus with $\mathrm{Aff}_{\mathsf C}(-\mathsf s_i)$ a torsor, $\psi_i$ is represented by an explicit weighted tropical cycle on the base, so algebraic $\psi$-class intersections can be read off from cone combinatorics.
- The genus-zero comparison is recovered: the tropicalization of $\mathsf{M}_{0,n}$ with the new affine structure is exactly the standard tropical space $\mathsf{M}^{\mathrm{trop}}_{0,n}$ with cross-ratio affine functions, so the theorem specializes to the known rational case.
- For families whose moduli map lands in the good locus $\mathsf{V}^{\mathrm{good}}_{g,n}$, the tropicalization is a family of tropical curves and the equality holds; this gives a practical criterion for when tropical $\psi$-class computations are trustworthy.
- The extended genus-one example shows the machinery works outside the rational case: tropical $\psi_1$ on families from admissible covers is the tropicalization of the algebraic class and agrees with operational tropicalization.
Reading between the lines
- (Pith inference) The same argument should carry over to products of $\psi$ classes and to tautological classes built from cotangent bundles, because the Chern-class map and tropicalization commute; the paper only states the single-class comparison.
- (Pith inference) The combinatorial-principality criterion gives a working definition of which algebraic divisor classes are visible to tropical geometry: a class is tropically visible exactly when its piecewise linear representative can be made affine on every stratum by subtracting an affine function.
- (Pith inference) The functoriality used in Section 6 suggests a practical recipe for higher-genus computations: pull affine functions back from genus-zero pieces of the tropicalization rather than constructing the affine sheaf from scratch, which may lower the cost of computing tautological intersections.
- (Pith inference) If the affine structure is invariant under log modifications, the comparison should continue to hold after blowing up the base or total space to remove self-intersections, so the theorem is likely applicable to semistable models as well as stable ones.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper develops a notion of tropicalization for toroidal embeddings that records not only the extended cone complex but also a sheaf of affine functions, defined by requiring that certain piecewise linear functions correspond to line bundles that trivialize on strata. This makes tropicalization a functor to the category of tropical spaces, invariant under logarithmic modifications. The main result is a comparison theorem: for a tropicalizable family of stable marked curves C → B whose tropicalization admits enough affine functions near the i-th section, the tropicalization of the algebraic cotangent line bundle L_i is isomorphic to the tropical cotangent line bundle L_trop_i of [CGM22], and consequently Trop(ψ_i) = ψ_trop_i. The paper also contains an extended example in genus one, computing tropical ψ classes on a space of admissible covers and verifying consistency with algebraic pushforward relations.
Significance. If the comparison theorem holds as stated, it resolves the motivating question of whether the combinatorial tropical ψ classes of [CGM22] are actual tropicalizations of algebraic ψ classes, under explicit and clearly stated hypotheses. The paper's main strengths are its careful conditional formulation, its functorial and modification-invariant definition of tropicalization, the detailed treatment of line bundles and affine structures, and a substantial worked example connecting the general theory to concrete computations in tropical admissible covers. The extended example is a genuine verification that is independent of the abstract formalism and adds credibility to the main claims. The paper does not rely on black-box software or unchecked computations; the arguments are presented in enough detail to be followed, with a small number of delegated 'immediate' checks.
major comments (1)
- [Corollary 5.11; Section 1.2] The statement of the headline comparison theorem omits a completeness/properness hypothesis on the base B, but its proof invokes Proposition 5.10, which begins 'Let X be a complete toroidal variety.' Definition 5.6 defines Trop_X(c) via the intersection number ∫_X c·[V(σ)], which is finite only when X is complete (or at least proper over the base field). Definition 3.18 of a tropicalizable family does not require B to be complete, and neither Corollary 5.11 nor the claim in Section 1.2 adds such a hypothesis. Thus for non-proper toroidal bases—for example an open subset of M_{g,n} or an affine base—the term Trop(ψ_i) on the left-hand side of Trop(ψ_i)=ψ_trop_i is not defined as a tropical cycle on the full extended cone complex. The paper's main examples use proper spaces and are unaffected, but the theorem as written is over-stated. Please add a completeness hypothesis to Corollary 5.11 and to the corresponding statement in the introduction and in Theorem F, or alternatively formulate and prove a local statement for proper families and explain precisely how the non-proper case is to be handled.
minor comments (4)
- [Proposition 3.16] The proof refers forward to Corollary 5.8 for the vanishing of the intersection product ϕ·[Σ^σ_X]. Since Corollary 5.8 is proved independently of Proposition 3.16, this is not circular, but the forward reference should be explicitly flagged (or the statements reordered) to avoid the appearance of a circular argument.
- [Example 4.3] In Cases 2 and 3, the text says the computations are 'very similar' and lists results without derivation. Since these examples are the main illustration of the difference between being a tropical line bundle on the cone complex and on the extended cone complex, it would be helpful to include the affine-function lists for the open sets U0y, Uxy, and U∞y in Case 2 and for U0y in Case 3, or to provide a table with the relevant triviality conditions.
- [Definition 5.6 and Corollary 5.11] The notation Trop(ψ_i) in Corollary 5.11 is not explicitly defined; once completeness of B is added as a hypothesis, the authors should state that Trop(ψ_i) means Trop_B(ψ_i) as in Definition 5.6, applied to the algebraic class c1(L_i) on B.
- [Remark 3.9] The claim that one may remove self-intersections 'by subdividing barycentrically the generalized cone complex' could use a brief clarification of why the subdivision can be chosen combinatorially, and why the resulting affine structure is independent of the choice by Proposition 3.8.
Circularity Check
No significant circularity: the comparison theorem is a genuine derivation from the paper's independently defined affine structure, and the CGM22 objects enter as definitions rather than as a closed loop.
full rationale
The paper's central claim, Theorem 5.5 / Corollary 5.11, is not circular. The affine structure on the tropicalization is defined in Definition 3.1 via triviality of the line bundle O_Xσ(φ) on strata, independently of the CGM22 psi class. Proposition 4.4 then gives a criterion for Trop(L) to be a tropical line bundle, and Proposition 5.2 identifies Trop(O_X(φ)) with the pseudo-torsor Aff_U(φ); this is a direct translation of the affine-structure definition, not an assumption of the conclusion. The target L_trop^i is introduced in Definition 5.4 'analogously to [CGM22, Definition 6.16]', but the theorem does not assume the equality Trop(L_i) ≅ L_trop^i; it derives it using Propositions 4.5, 5.2 and 5.3. No parameters are fitted, no uniqueness theorem is imported to forbid alternatives, and no known result is merely renamed. Self-citations to CGM22 occur for definitions and for supporting results such as the cross-ratio description of affine functions on M_trop_0,n and the example's use of [CGM22, Prop 7.8]; these are published, parameter-free results with stated assumptions, so they function as independent evidence rather than a circular chain. One non-circular correctness gap should be flagged separately: Corollary 5.11 applies Proposition 5.10, whose statement begins 'Let X be a complete toroidal variety', while Definition 3.18 and the corollary do not require completeness of B; thus Trop(ψ_i) is not defined for non-proper bases as written. This is an over-statement in the theorem, not a reduction of the conclusion to its inputs. Overall, the derivation is self-contained and the headline equality is a substantive comparison, not a definitional identity.
Assumptions & free parameters
assumptions (4)
- domain assumption The base field is algebraically closed of characteristic 0.
- domain assumption The definitions and results on families of tropical curves, tropical psi classes, and tropical cycles from [CGM22] are taken as given.
- standard math Toroidal embeddings without self-intersections can be assumed after suitable log modifications, and the affine structure is invariant under such modifications.
- ad hoc to paper The family under consideration is tropicalizable, i.e., satisfies Definition 3.18, including condition (3) on irreducibility of generic fibers.
Cite this review
Pith. "Pith review of Tropicalization of $\psi$ classes." pith.science (2026). https://pith.science/paper/ADRPQSVV
@misc{pith2026241202817,
author = {Pith},
title = {Pith review of: Tropicalization of $\psi$ classes},
year = {2026},
howpublished = {\url{https://pith.science/paper/ADRPQSVV}},
note = {Machine review of arXiv:2412.02817}
}
abstract
Under suitable conditions on a family of logarithmic curves, we endow the tropicalization of the family with an affine structure in a neighborhood of the sections in such a way that the tropical $\psi$ classes from \cite{psi-classes} arise as tropicalizations of algebraic $\psi$ classes.
Forward citations
Cited by 1 Pith paper
-
Pure extension of the theta divisor over the moduli space of abelian varieties
The pure weight-2 extension of the universal theta divisor is the Zariski closure twisted by div θinv, where θinv is the difference of the tropical and smooth tropical Riemann theta functions; Moret-Bailly's key formu...
Reference graph
Works this paper leans on
-
[1]
The tropicalization of the moduli space of curves
Dan Abramovich, Lucia Caporaso, and Sam Payne. The tropicalization of the moduli space of curves. Ann. Sci. Ec. Norm. Super. , 48(4):765--809, 2015
work page 2015
-
[2]
Billera, Susan P
Louis J. Billera, Susan P. Holmes, and Karen Vogtmann. Geometry of the space of phylogenetic trees. Adv. in Appl. Math. , 27(4):733--767, 2001
2001
-
[3]
Gonality of algebraic curves and graphs
Lucia Caporaso. Gonality of algebraic curves and graphs. In Algebraic and complex geometry , volume 71 of Springer Proc. Math. Stat. , pages 77--108. Springer, Cham, 2014
work page 2014
-
[4]
Dustin Cartwright. Tropical complexes. Manuscr. Math. , 163(1-2):227--261, 2020
work page 2020
-
[5]
A moduli stack of tropical curves
Renzo Cavalieri, Melody Chan, Martin Ulirsch, and Jonathan Wise. A moduli stack of tropical curves. 2017
work page 2017
-
[6]
Renzo Cavalieri, Andreas Gross, and Hannah Markwig. Tropical \( \) classes. Geom. Topol. , 26(8):3421--3524, 2022
work page 2022
-
[7]
Topology of moduli spaces of tropical curves with marked points
Melody Chan, S ren Galatius, and Sam Payne. Topology of moduli spaces of tropical curves with marked points. In Facets of algebraic geometry. A collection in honor of William Fulton's 80th birthday. Volume 1 , pages 77--131. Cambridge: Cambridge University Press, 2022
work page 2022
-
[8]
Tropicalizing the space of admissible covers
Renzo Cavalieri, Hannah Markwig, and Dhruv Ranganathan. Tropicalizing the space of admissible covers. Math. Ann. , 364(3-4):1275--1313, 2016
2016
Show all 25 references
-
[9]
Cocycles on tropical varieties via piecewise polynomials
Georges Francois. Cocycles on tropical varieties via piecewise polynomials. Proc. Am. Math. Soc. , 141(2):481--497, 2013
2013
-
[10]
Matroid polytopes, nested sets and Bergman fans
Eva Maria Feichtner and Bernd Sturmfels. Matroid polytopes, nested sets and Bergman fans. Port. Math. (N.S.) , 62(4):437--468, 2005
2005
-
[11]
Forms on berkovich spaces based on harmonic tropicalizations
Walter Gubler, Philipp Jell, and Joe Rabinoff. Forms on berkovich spaces based on harmonic tropicalizations. Preprint: arXiv:2111.05741 , 2021
2021 arXiv
-
[12]
Logarithmic intersections of double ramification cycles
David Holmes and Rosa Schwarz. Logarithmic intersections of double ramification cycles. Algebr. Geom. , 9(5):574--605, 2022
2022
-
[13]
Topological recursive relations in H^ 2g ( M_ g,n )
Eleny-Nicoleta Ionel. Topological recursive relations in H^ 2g ( M_ g,n ) . Invent. Math. , 148(3):627--658, 2002
2002
-
[14]
Lefschetz (1,1) -theorem in tropical geometry
Philipp Jell, Johannes Rau, and Kristin Shaw. Lefschetz (1,1) -theorem in tropical geometry. \' E pijournal Geom. Alg\' e brique , 2:Art. 11, 27, 2018
2018
-
[15]
Tropical toric geometry
Takeshi Kajiwara. Tropical toric geometry. In Toric topology , volume 460 of Contemp. Math. , pages 197--207. Amer. Math. Soc., Providence, RI, 2008
2008
-
[16]
Tropical intersection theory from toric varieties
Eric Katz. Tropical intersection theory from toric varieties. Collect. Math. , 63(1):29--44, 2012
2012
-
[17]
Kempf, Finn Faye Knudsen, D
G. Kempf, Finn Faye Knudsen, D. Mumford, and B. Saint-Donat. Toroidal embeddings. I , volume Vol. 339 of Lecture Notes in Mathematics . Springer-Verlag, Berlin-New York, 1973
1973
-
[18]
Notes on psi classes
Joachim Kock. Notes on psi classes. Notes. http://mat.uab.es/ kock/GW/notes/psi-notes.pdf , 2001
2001
-
[19]
okova G eometry- T opology C onference 2006 , pages 39--51. G\
Grigory Mikhalkin. Moduli spaces of rational tropical curves. In Proceedings of G \"okova G eometry- T opology C onference 2006 , pages 39--51. G\"okova Geometry/Topology Conference (GGT), G\"okova, 2007
2006
-
[20]
Molcho, R
S. Molcho, R. Pandharipande, and J. Schmitt. The H odge bundle, the universal 0-section, and the log C how ring of the moduli space of curves. Compos. Math. , 159(2):306--354, 2023
2023
-
[21]
A case study of intersections on blowups of the moduli of curves
Sam Molcho and Dhruv Ranganathan. A case study of intersections on blowups of the moduli of curves. Algebra Number Theory , 18(10):1767--1816, 2024
2024
-
[22]
Introduction to tropical geometry , volume 161 of Graduate Studies in Mathematics
Diane Maclagan and Bernd Sturmfels. Introduction to tropical geometry , volume 161 of Graduate Studies in Mathematics . American Mathematical Society, Providence, RI, 2015
2015
-
[23]
Tropical curves, their J acobians and theta functions
Grigory Mikhalkin and Ilia Zharkov. Tropical curves, their J acobians and theta functions. In Curves and abelian varieties , volume 465 of Contemp. Math. , pages 203--230. Amer. Math. Soc., Providence, RI, 2008
2008
-
[24]
Functorial tropicalization of logarithmic schemes: the case of constant coefficients
Martin Ulirsch. Functorial tropicalization of logarithmic schemes: the case of constant coefficients. Proc. Lond. Math. Soc. (3) , 114(6):1081--1113, 2017
2017
-
[25]
Non-archimedean geometry of artin fans
Martin Ulirsch. Non-archimedean geometry of artin fans. Advances in Mathematics , 345:346--381, 2019
2019
Reviewed August 11, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.