← All problems

Hodge conjecture

76 years old · traced to

Let XX be a non-singular complex projective variety of complex dimension nn. Its singular cohomology carries the Hodge decomposition

Hm(X,C)=Hm(X,Q)⊗QC=⨁p+q=mHp,q(X),Hp,q(X)‾=Hq,p(X),H^m(X,\mathbb C)=H^m(X,\mathbb Q)\otimes_{\mathbb Q}\mathbb C=\bigoplus_{p+q=m}H^{p,q}(X),\qquad \overline{H^{p,q}(X)}=H^{q,p}(X),

where Hp,q(X)H^{p,q}(X) is the space of classes represented by closed forms of type (p,q)(p,q), that is, forms that in local holomorphic coordinates are sums of terms a dzi1∧⋯∧dzip∧dzˉj1∧⋯∧dzˉjqa\,dz_{i_1}\wedge\cdots\wedge dz_{i_p}\wedge d\bar z_{j_1}\wedge\cdots\wedge d\bar z_{j_q}. For 0≤k≤n0\le k\le n, the space of Hodge classes of degree 2k2k is

Hdg⁡k(X)=H2k(X,Q)∩Hk,k(X).\operatorname{Hdg}^k(X)=H^{2k}(X,\mathbb Q)\cap H^{k,k}(X).

Every closed irreducible algebraic subvariety Z⊆XZ\subseteq X of codimension kk has a fundamental class cl⁡(Z)∈H2k(X,Q)\operatorname{cl}(Z)\in H^{2k}(X,\mathbb Q), Poincaré dual to the homology class of ZZ. Integration over ZZ shows that cl⁡(Z)\operatorname{cl}(Z) has type (k,k)(k,k), so cl⁡(Z)∈Hdg⁡k(X)\operatorname{cl}(Z)\in\operatorname{Hdg}^k(X). Let Ak(X)⊆Hdg⁡k(X)A^k(X)\subseteq\operatorname{Hdg}^k(X) be the Q\mathbb Q-span of all such classes.

Hodge conjecture. For every non-singular complex projective variety XX and every kk,

Hdg⁡k(X)=Ak(X).\operatorname{Hdg}^k(X)=A^k(X).

That is, every rational cohomology class of type (k,k)(k,k) on XX is a Q\mathbb Q-linear combination of classes of algebraic subvarieties of codimension kk.

This is the statement in Deligne's official Clay problem description [1]. By Chow's theorem, algebraic subvarieties may be replaced by closed analytic subvarieties. The cases k=0k=0, k=1k=1 and k=nk=n are theorems, and by hard Lefschetz so is k=n−1k=n-1. A solution must prove the statement for all XX and all kk, or exhibit a single XX and a Hodge class not in Ak(X)A^k(X).

Background

Hodge raised the question at the 1950 International Congress [2], in an integral form that turned out to be false. The rational version stated above is the one that survives. It predicts that the purely topological and analytic notion of a Hodge class detects algebraic cycles exactly. Over other fields the analogous prediction is the Tate conjecture for ℓ\ell-adic cohomology.

The conjecture sits at the center of the theory of motives. If the Künneth components of the diagonal and the inverse of the Lefschetz operator are algebraic, Grothendieck's motives over C\mathbb C form a semisimple Tannakian category. The full Hodge conjecture is then equivalent to the Hodge realization functor being fully faithful. It also implies that the locus where a class becomes Hodge in a family is algebraic and defined over Qˉ\bar{\mathbb Q}. [1, 7]

Equivalent formulations
  1. In mathematics, the Hodge conjecture is a major unsolved problem in algebraic geometry and complex geometry that relates the algebraic topology of a non-singular complex algebraic variety to its subvarieties.

    source: Wikipedia

References
  1. P. Deligne, "The Hodge conjecture", in J. Carlson, A. Jaffe and A. Wiles (eds.), The Millennium Prize Problems, Clay Math. Inst. and AMS (2006). Clay PDF
  2. W. V. D. Hodge, "The topological invariants of algebraic varieties", in Proc. International Congress of Mathematicians (Cambridge, Mass., 1950), Vol. 1, AMS (1952), 182–192.
  3. K. Kodaira and D. C. Spencer, "Divisor class groups on algebraic varieties", Proc. Natl. Acad. Sci. USA 39 (1953), 872–877. doi:10.1073/pnas.39.8.872
  4. A. Conte and J. P. Murre, "The Hodge conjecture for fourfolds admitting a covering by rational curves", Math. Ann. 238 (1978), 79–88. doi:10.1007/BF01351457
  5. T. Shioda, "The Hodge conjecture for Fermat varieties", Math. Ann. 245 (1979), 175–184. doi:10.1007/BF01428804
  6. B. J. J. Moonen and Yu. G. Zarhin, "Hodge classes on abelian varieties of low dimension", Math. Ann. 315 (1999), 711–733. doi:10.1007/s002080050333
  7. C. Voisin, "The Hodge conjecture", in Open Problems in Mathematics, Springer (2016), 521–543. doi:10.1007/978-3-319-32162-2_17
  8. E. Markman, "The monodromy of generalized Kummer varieties and algebraic cycles on their intermediate Jacobians", J. Eur. Math. Soc. 25 (2023), 231–321. doi:10.4171/JEMS/1199
  9. E. Markman, "Cycles on abelian 2n2n-folds of Weil type from secant sheaves on abelian nn-folds", preprint (2025). arXiv:2502.03415
  10. E. Cattani, P. Deligne and A. Kaplan, "On the locus of Hodge classes", J. Amer. Math. Soc. 8 (1995), 483–506. doi:10.1090/S0894-0347-1995-1273413-2
  11. M. F. Atiyah and F. Hirzebruch, "Analytic cycles on complex manifolds", Topology 1 (1962), 25–45. doi:10.1016/0040-9383(62)90094-0
  12. C. Voisin, "A counterexample to the Hodge conjecture extended to Kähler varieties", Int. Math. Res. Not. 2002, no. 20, 1057–1075. doi:10.1155/S1073792802111135
  13. Wikipedia, Hodge conjecture.

Source: Wikipedia

Progress summary

Refreshed
Open · progress reported

The full conjecture is still open, while a newly public manuscript claims progress for a narrow family of varieties.

Proposed by W. V. D. Hodge in 1950, the conjecture asserts that every rational cohomology class of type (k,k)(k,k) is represented by algebraic cycles. A 2024 expert interview described substantial special-case knowledge but no general proof; a later report mentions an unverified proof of one particular case.

Known results

  • Codimension 11 is settled by the Lefschetz theorem on classes of type (1,1)(1,1).
  • The conjecture is known in various low-dimensional and specially structured cases.
  • The integral analogue is false (Atiyah–Hirzebruch, 1962).

Community submission (unverified)

Posted October 7, 2026: A submitted claim attributes the rational Hodge conjecture for CM abelian varieties to OpenAI and cites a September 30 manuscript. This concerns a special family, not the general conjecture, and its mathematical correctness is unverified.

Current status (as of October 2026): The general Hodge conjecture remains open; reported progress for CM abelian varieties is unverified.

Sources

Solutions 2

Claimed by OpenAI. The manuscript claims the rational Hodge conjecture in every codimension for all finite products of projective complex K3 surfaces, including repeated factors and arbitrary periods and Picard numbers. The claim concerns this family of varieties and does not settle the general Hodge conjecture.See full solutionHide full solution

Claimed by OpenAI.

The manuscript claims the rational Hodge conjecture in every codimension on every finite product of projective complex K3 surfaces, with distinct or repeated factors and no restrictions on their periods, Picard numbers or endomorphism fields. It does not claim the Hodge conjecture for arbitrary smooth projective varieties.

GitHub repository: https://github.com/openai/math

Manuscript: https://github.com/openai/math/blob/adc7f1241b42e322a6451854ab7e4b4c146bf78a/preprints/The-rational-Hodge-conjecture-for-products-of-K3-surfaces-October-4-2026/hodge-conjecture-products-k3.pdf

  • hodge-conjecture-products-k3 (1).pdf639,139 bytesOpen
1 reply to this solution
Claimed by OpenAI. The manuscript claims the rational Hodge conjecture for complex abelian varieties of CM type: every rational Hodge class is represented by a rational linear combination of algebraic cycle classes. The result is stated for CM abelian varieties, a special case of the general Hodge conjecture.See full solutionHide full solution

Claimed by OpenAI. The attached preprint concerns the rational Hodge conjecture for CM abelian varieties, a special case of the general Hodge conjecture.

GitHub repository: https://github.com/openai/math/tree/main

Original PDF: https://github.com/openai/math/blob/main/preprints/The-rational-Hodge-conjecture-for-CM-abelian-varieties-September-30-2026/paper.pdf

  • 032.pdf427,044 bytesOpen