← All problems

Jacobian conjecture and Hessian conjecture in the gradient case

Does every polynomial h∈C[X1,X2,X3,X4]h\in\mathbb{C}[X_1,X_2,X_3,X_4] with constant nonzero Hessian determinant, det⁡(∂2h∂Xi∂Xj)∈C∗\det\left(\frac{\partial^2h}{\partial X_i\partial X_j}\right)\in\mathbb{C}^{*}, have a polynomially invertible gradient map ∇h=(∂1h,∂2h,∂3h,∂4h):C4→C4\nabla h=(\partial_1h,\partial_2h,\partial_3h,\partial_4h):\mathbb{C}^{4}\to\mathbb{C}^{4}?

Equivalent formulations
  1. For a polynomial h∈C[X1,X2,X3,X4]h\in\mathbb{C}[X_1,X_2,X_3,X_4] with constant nonzero Hessian determinant, set g=∑i=14Xi∂ih−hg=\sum_{i=1}^{4}X_i\partial_i h-h. Then ∇h\nabla h is a polynomial automorphism if and only if g∈C[∂1h,∂2h,∂3h,∂4h]g\in\mathbb{C}[\partial_1h,\partial_2h,\partial_3h,\partial_4h].

    source: Adamus and Hajto, “The Legendre potential and the gradient case of the differential Galois criterion for the Jacobian Conjecture”

References

Progress summary

Refreshed
Open · progress reported

The gradient version is settled through three dimensions, while a claimed five-dimensional counterexample and a claimed four-dimensional quartic result leave the key four-dimensional and planar cases open.

The problem concerns the relationship between the Jacobian and Hessian conjectures for polynomial maps that are gradients. In this setting, the Hessian formulation is equivalent to the Jacobian conjecture, but the remaining cases are not resolved.

Known results

  • Meng, 2003 (revised 2005): the Hessian conjecture is equivalent to the Jacobian conjecture and holds over R\mathbb{R} under a definite-Hessian hypothesis.
  • Gradient-form Jacobian conjecture: proved for n≤3n\leq 3 over characteristic-zero fields.
  • The four-dimensional Hessian conjecture HC4\mathrm{HC}_{4} implies the planar Jacobian conjecture JC2\mathrm{JC}_{2}.

2026 claimed dimension-five counterexample and dimension-four advance

Meng and Yang claim an explicit degree-1414 polynomial disproves the Hessian conjecture for n≥5n\geq 5, leaving only HC4\mathrm{HC}_{4} and JC2\mathrm{JC}_{2} unsettled. Separately, Ni claims the quartic case of HC4\mathrm{HC}_{4}, while explicitly leaving degrees at least five open. Adamus and Hajto give a single-condition criterion and inverse-recovery framework for the gradient case. These claims have no independent verification in the retrieved sources.

Current status (as of October 2026): the gradient case is established for n≤3n\leq 3; the claimed result for n≥5n\geq 5 and quartic result in n=4n=4 are unverified, while HC4\mathrm{HC}_{4} and JC2\mathrm{JC}_{2} remain open.

Sources

Solutions 0

No solutions have been posted yet.