Pith. sign in

REVIEW 4 major objections 4 minor 21 references

A remark on algebraic cycles on cubic fourfolds

T0 review · 4 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read The paper claims that $A_1$ of a smooth cubic fourfold is the kernel of a push-forward between two surfaces, and derives a non-rationality criterion from this structure.

desk verdict Interesting idea—controlling A1 of a cubic fourfold via A0 of a double cover—but the central criterion rests on a false blow-up formula and the two descriptions of A1 are never reconciled. read the letter →

arxiv 1908.04576 v1 pith:JVHFTTAB submitted 2019-08-13 math.AG math.KT

Kalyan Banerjee This is my paper · ORCID
classification math.AGmath.KT MSC 14C2514J3514M20
keywords algebraiccyclescubicfourfoldChowgroupsrationalityone-cycleszero-cyclescorrespondencesbirationalinvariance
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 proposes to generalize the classical non-rationality criterion for smooth cubic threefolds to smooth cubic fourfolds by replacing the intermediate Jacobian with the group $A_1$ of algebraically trivial one-cycles modulo rational equivalence. It claims that $A_1$ of any smooth cubic fourfold is isomorphic to the kernel of a push-forward between two smooth projective surfaces attached to the fourfold, and that it is not dominated by the zero-cycle group of any single surface. On this basis it defines weak representability up to dimension 2 and claims that rationality forces $A_1$ to be weakly representable up to dimension 2, with the kernel of the domination map a finite sum of zero-cycle groups of curves and surfaces. If that kernel is not such a finite sum, the paper concludes the cubic fourfold is non-rational. It also derives that a cubic fourfold birational to projective four-space cannot be obtained by a single blow-up followed by a single blow-down.

What carries the argument

The mechanism is the conic-bundle presentation of a cubic fourfold obtained by projecting from a line $L$: the discriminant surface $S$ parametrizes reducible conics, $\widetilde S$ is the double cover in the Fano variety of lines, and the universal family gives a correspondence $Z$ inducing a surjection $A_0(\widetilde S)\to A_1(X)$. The kernel of $Z_*$ is identified as exactly $\pi^*A_0(S)$, using monodromy and the Picard-Lefschetz formula to show that any abelian subvariety in the kernel of $J(\widetilde S_t)\to A_1(X_t)$ lies in $\pi^*J(S_t)$. This yields the isomorphism $A_1(X)\cong \ker(A_0(\widetilde S)\to A_0(S))$. The final section introduces weak representability up to dimension 2 and relies on a decomposition of $A_1$ after blowing up an indeterminacy locus—$A_1(\widetilde X)\cong A_1(X)\oplus J(C)$ or $A_1(X)\oplus A_0(S)\oplus A_1(S)$—to claim birational invariance.

What would settle it

Test the splitting formula directly: blow up $\mathbb P^4$ along a smooth surface $S$ with $h^{2,0}(S)>0$ and compute $A_1$ of the blow-up; if it is not $A_1(\mathbb P^4)\oplus A_0(S)\oplus A_1(S)$, the decomposition used in Theorem 6.3 is false and the rationality criterion loses its basis.

Watch

Extended reading notes

Core claim

The central claim is that for a smooth cubic fourfold $X$, the group $A_1(X)$—one-cycles algebraically equivalent to zero modulo rational equivalence—is isomorphic to the kernel of a push-forward $\pi_*: A_0(\widetilde S)\to A_0(S)$, where $S$ is the discriminant surface of the projection from a line and $\widetilde S$ is its double cover inside the Fano variety of lines. Equivalently, $A_1(X)\cong A_0(\widetilde S)/\pi^*A_0(S)$. The paper further claims this group is not dominated by $A_0(S')$ for any single smooth projective surface $S'$, and it introduces weak representability up to dimension 2: $A_1(X)$ is weakly representable up to dimension 2 when it is dominated by a finite sum of $A_0$'s of curves and surfaces. The paper claims that rationality forces this weak representability and forces the kernel of the domination map to be a finite sum of $A_0$'s of curves and surfaces; therefore a cubic whose kernel fails that condition must be non-rational. It also claims birational invariance of weak representability up to dimension 2, and uses this to prove that a cubic fourfold birational to $\mathbb P^4$ cannot be a single blow-up followed by a single blow-down.

Load-bearing premise

The load-bearing premise is the unproved splitting formula for one-cycles after blowing up the indeterminacy locus of a birational map; if that splitting is wrong, the birational-invariance theorem and the non-rationality criterion both collapse.

Editorial extensions

If this is right

  • If a smooth cubic fourfold is rational, then $A_1(X)$ is weakly representable up to dimension 2, dominated by finitely many zero-cycle groups of curves and surfaces.
  • If the kernel of the map from a finite sum of such zero-cycle groups to $A_1(X)$ is not itself a finite sum of zero-cycle groups of curves and surfaces, the cubic fourfold is non-rational.
  • No smooth cubic fourfold has $A_1(X)$ isomorphic to $A_0(S')$ for a single smooth projective surface $S'$.
  • A cubic fourfold birational to $\mathbb P^4$ cannot be obtained by blowing up $\mathbb P^4$ once along a curve or surface and then blowing down once.
  • For a cubic containing a plane whose associated quadric bundle is rational, the quadric bundle must be produced by more than one blow-up followed by blow-downs, while its $A_1$ is still weakly representable up to dimension 2.

Reading between the lines

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

  • Extension: If the splitting formula in Theorem 6.3 is supplied with a proof, the criterion becomes a concrete test for non-rationality: compute the kernel of $A_0(\widetilde S)\to A_0(S)$ and check whether it is a finite sum of $A_0$'s of curves and surfaces.
  • Extension: The criterion suggests a measure of how far a cubic fourfold is from being rational, namely the length of a birational chain to $\mathbb P^4$, and predicts rational examples must lie in a countable union of special loci in the moduli space of cubic fourfolds.
  • Extension: The rational examples of cubic fourfolds built from quadric bundles should satisfy the kernel splitting condition, which would show that the new criterion is not vacuous.
  • Extension: For a very general cubic fourfold, if the kernel is not a finite sum of zero-cycle groups of curves and surfaces, the criterion would certify non-rationality and would align with the Chow-theoretic decomposition-of-the-diagonal approach to stable rationality.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 4 minor

Summary. The paper proposes to generalize the Clemens-Griffiths non-rationality criterion from cubic threefolds to cubic fourfolds by studying A1(X), the group of algebraically trivial one-cycles modulo rational equivalence. The author constructs, from a line on a smooth cubic fourfold X, a conic bundle structure and a discriminant surface S with a double cover ~S, and claims that A0(~S) surjects onto A1(X). The main structural claim is that A1(X) is isomorphic to A0(~S)/π^*A0(S), and later the paper asserts instead that A1(X) is the kernel of the push-forward π_*: A0(~S) -> A0(S). Based on this, the paper defines weak representability of A1 up to dimension 2 and claims that it is a birational invariant, leading to a proposed non-rationality criterion for cubic fourfolds. The introduction announces a main result but the announcement is followed by an open question rather than a theorem statement.

Significance. If the main claims were correct, the paper would offer a new structural description of A1 for cubic fourfolds and a concrete rationality criterion in terms of zero-cycles on surfaces. The idea of using a correspondence between two surfaces to control A1 is potentially interesting, and the paper correctly draws on Schoen's theorem and cites Voisin's independent approach. However, the paper's central results are not established: the quotient and kernel descriptions of A1 are never identified, the birational-invariance theorem rests on an unsupported and apparently incorrect blow-up formula, and the proof of the main structural theorem has a gap in passing from abelian subvarieties to the full kernel. These are load-bearing issues, not presentation issues.

major comments (4)
  1. [§5, Theorem 5.1 and Remark 5.2] Theorem 5.1 states that A1(X) is isomorphic to the quotient A0(~S)/π^*A0(S), whereas Remark 5.2 and the beginning of Section 6 state that A1(X) is the kernel of the push-forward π_*: A0(~S) -> A0(S). These are not identified in the manuscript, and for a degree-2 finite morphism a quotient by a pullback subgroup is not generally isomorphic to the kernel of the push-forward. Since Section 6 uses the kernel description as the basis for Theorem 6.1 and the subsequent non-representability claims, the main structural statement is ambiguous and the later arguments are built on an unproved identification.
  2. [§6, Theorem 6.3] The proof of Theorem 6.3 asserts, in one sentence, that after blowing up the indeterminacy locus of a birational map of fourfolds, A1(~X) is isomorphic to A1(X)⊕J(C) or to A1(X)⊕A0(S)⊕A1(S). This is inconsistent with the standard blow-up formula: for a smooth center Z of codimension c, A_1(Bl_Z X) is isomorphic to A_1(X) ⊕ ⊕_{i=1}^{c-1} A_{1-i}(Z), so a surface center gives A_1(X)⊕A_0(Z) and a curve center gives A_1(X)⊕A_0(Z), not the terms asserted in the paper. The birational invariance of weak representability up to dimension 2, and the rationality criterion that depends on it, therefore have no valid supporting argument.
  3. [§5, proof of Theorem 5.1] The proof of Theorem 5.1 shows that each abelian subvariety A_i arising in the kernel of the composed map lies in π_i^*J(S_{t_i}), but the kernel had been identified only as a countable union of shifts of an abelian variety. The passage from 'each A_i is contained in π^*J(S_t)' to 'every element of the kernel of Z_* lies in π^*A0(S)' is not justified, because no argument controls the translation components of the countable union. This gap is load-bearing, since the conclusion that A0(~S)/π^*A0(S) is isomorphic to A1(X) depends on it.
  4. [§1 and abstract] The announced 'main result of this paper' is never stated as a theorem: the text says 'The main result of this paper is as follows:' and then immediately poses a question about very general cubic fourfolds. The abstract also says the paper 'tries to generalize' the Clemens-Griffiths criterion, which does not specify the claimed theorem. As a result, the reader cannot verify what the paper's central claim is.
minor comments (4)
  1. [§1] There is a typo 'the the kernel' in the paragraph following the displayed commutative diagram.
  2. [§2 and §5] The notation is inconsistent: the surfaces are called S, T, and ~S in different places, and the correspondence maps are sometimes written with subscripts that are not defined. This makes the proofs harder to follow.
  3. [§2] The term 'essential dimension 2' is used in the introduction but is not formally defined; the later definition of weak representability up to dimension 2 appears only in Definition 6.2.
  4. [References] The key lemma about countable unions of Zariski closed subsets is cited to [BG], the author's own prior arXiv paper, but the precise statement is not reproduced; the reader is forced to consult an unpublished source for a load-bearing step.

Circularity Check

0 steps flagged · score 2.0 of 10

No circular reduction found: the main non-representability inputs are external (Schoen, Mumford), and the self-citation [BG] is a technical lemma rather than a re-definition of the target result.

full rationale

The paper's derivation chain does not contain a step in which a claimed prediction is equal by construction to a fitted input or to a self-citation containing the target theorem. Theorem 2.1 uses an explicit conic-bundle construction and cites Schoen for the non-domination by a Jacobian; this is an external benchmark. The countable-union/abelian-subvariety lemma in Theorem 2.2 is attributed to the author's own [BG]; although this is a self-citation and is used in the later kernel identification, it is a general lemma about kernels of maps between Chow groups and not itself the assertion that A1(X) is a kernel between two surfaces, so the central claim does not reduce to the citation. Theorem 5.1 proves A1(X) is isomorphic to A0(~S)/π^*A0(S), and Remark 5.2 asserts without proof that A1(X) is the kernel of the push-forward A0(~S)→A0(S); these are different objects, and the latter is an unsupported identification rather than a circular derivation. Similarly, Theorem 6.3's blow-up formula for A1(~X) is asserted in one sentence and appears inconsistent with the standard blow-up formula; this is a serious correctness gap but not a circularity, because no equation in the paper shows the output equals its own input. No parameters are fitted to data and no empirical quantity is renamed as a prediction. Accordingly the circularity score is low.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

The proof relies on several unproved background results. The central inputs are Schoen's non-representability of A1(X), a subgroup-structure lemma from the author's own [BG], a constant-dimension monodromy identification from [Be], and an asserted blow-up decomposition. None of these is derived in the note.

assumptions (4)
  • domain assumption A1(X) is not representable: it is not dominated by the Jacobian of a curve or by A0 of a single surface.
    Used in Section 2 (Theorem 2.2) and Section 5 to exclude the alternative that the monodromy-stable abelian subvariety maps surjectively onto A1(X_t); attributed to Schoen [Sc] but not proved here.
  • ad hoc to paper If a subgroup of A0(T_t) is a countable union of Zariski closed subsets, then it is a countable union of shifts of an abelian subvariety inside J(T_t).
    Stated in Section 2 with reference to the author's own [BG]; this lemma is the engine for the kernel description in Theorems 2.2 and 5.1 and is not proved in this note.
  • domain assumption For the double cover ~S -> S, the kernel of J(~S_t) -> A1(X_t) has constant dimension and equals pi^*(J(S_t)) for general t.
    Used in Theorem 5.1's conclusion that every element of the kernel of Z_* lies in pi^*(A0(S)); the constancy argument depends on the map being a map of pi_1(U,t)-modules and on the special-case computation from [Be].
  • ad hoc to paper Blowing up the indeterminacy locus of a birational map of fourfolds gives A1(~X) = A1(X)⊕J(C) or A1(X)⊕A0(S)⊕A1(S).
    Assumed in Theorem 6.3 to prove birational invariance of weak representability up to dimension 2; no proof or citation is supplied.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A remark on algebraic cycles on cubic fourfolds." pith.science (2026). https://pith.science/paper/JVHFTTAB

@misc{pith2026190804576,
  author       = {Pith},
  title        = {Pith review of: A remark on algebraic cycles on cubic fourfolds},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JVHFTTAB}},
  note         = {Machine review of arXiv:1908.04576}
}
read the original abstract

In this short note we try to generalize the Clemens-Griffiths criterion of non-rationality for smooth cubic threefolds to the case of smooth cubic fourfolds.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

21 extracted references · 15 canonical work pages

  1. [1]

    N.Addington, B.Hassett, Y.Tschinkel, A.Varilly-Alvarado, Cubic fourfolds fibered into sextic del pezzo surfaces , https://arxiv.org/pdf/1606.05321.pdf, 2016

    work page Pith review arXiv 2016

  2. [2]

    Y.Andre, B.Kahn, (with appendix by P.O'Sullivan) Nilpotence, radicaux et structure monoidales , Rend. Sem. Math. Univ. Padova 108(2002) 107-291

    work page 2002

  3. [3]

    (French) , Ann

    A.Beauville, Variétés de Prym et jacobiennes intermédiaires. (French) , Ann. Sci. École Norm. Sup. (4) 10 (1977), no. 3, 309–391

    work page 1977

  4. [4]

    Banerjee and V

    K. Banerjee and V. Guletskii, Rational equivalence for line configurations on cubic hypersurfaces in P^5 . , arXiv:1405.6430v1 , 2014

    arXiv 2014

  5. [5]

    S.Bloch, Lectures on algebraic cycles , Duke University, Mathematics series IV

    work page

  6. [6]

    H.Clemens, P.Griffiths, The Intermediate Jacobian of the cubic threefold , Annals of Math., second series, vol. 95, no. 2, 1972, 281-356

    1972

  7. [7]

    (French) [Quartic hypersurfaces in dimension 3: stable non-rationality] , Ann

    J.Colliot-Thelene, A.Pirutka, Hypersurfaces quartiques de dimension 3: non-rationalité stable. (French) [Quartic hypersurfaces in dimension 3: stable non-rationality] , Ann. Sci. Éc. Norm. Supér. (4) 49 (2016), no. 2, 371–397

    work page 2016

  8. [8]

    900, Springer, 1982, 101-208

    P.Deligne, J.Milne, Tannakian categories , In Hodge cycles and Shimura varieties, Lecture notes in Math. 900, Springer, 1982, 101-208

    work page 1982

Show all 21 references
  1. [9]

    V.Guletskii, On the continuous part of codimension two algebraic cycles on threefolds over a field , Math Sbornik, Volume 200, number 3

  2. [10]

    V.Guletskii, C.Pedrini, The Chow motive of the Godeaux surface , Algebraic Geometry, a volume in the memory of Paolo Francia, Walter De Gruyter, 2002, 179-195

    2002

  3. [11]

    B.Hassett, Special cubic fourfolds , Compositio Math., Vol 120, no. 1, 1-23

  4. [12]

    B.Hassett, Some rational cubic fourfolds , Journal of Algebraic Geometry, 8, 1999, no.1, 103-114

    1999

  5. [13]

    S.Kimura, Chow groups are finite dimensional in some sense , Math.Ann., 331 (2005), no.1, 173-201

    2005

  6. [14]

    , J.Math Kyoto Univ

    D.Mumford, Rational equivalence for 0 -cycles on surfaces. , J.Math Kyoto Univ. 9, 1968, 195-204

    1968

  7. [15]

    A.Roitman, -equivalence of zero dimensional cycles (Russian) , Math. Sbornik. 86(128), 1971, 557-570

    1971

  8. [16]

    88, 1993

    C.Schoen, On Hodge structures and non-representability of Chow groups , Composition Math. 88, 1993

    1993

  9. [17]

    A.Suslin, V.Voevodsky Relative cycles and Chow sheaves , AM-143, Princeton Univertsity Press, 2000

    2000

  10. [18]

    Voisin, Complex algebraic geometry and Hodge theory II , Cambridge University Press, Cambridge, 2003

    C. Voisin, Complex algebraic geometry and Hodge theory II , Cambridge University Press, Cambridge, 2003

    2003

  11. [19]

    C.Voisin, Symplectic invoultions of K 3 surfaces act trivially on CH_0 , Documenta Mathematicae 17, 851-860, 2012

    2012

  12. [20]

    Society, 19, 1619-1653, 2014

    C.Voisin, Universal CH_0 of cubic hypersurfaces , Journal of European Math. Society, 19, 1619-1653, 2014

    2014

  13. [21]

    201, 207-237, 2015

    C.Voisin, Unirational threefolds with no universal codimension 2 cycle , Inventiones Math. 201, 207-237, 2015

    2015

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.