New complexity bounds for decomposing algebraic varieties

Researchers improved complexity bounds from PSPACE to the counting hierarchy CH for several problems.

Decomposing an algebraic variety into structured components is a fundamental problem. These components can be equidimensional, irreducible, or connected.

The work studies the computational complexity of computing equidimensional components using Chow forms. The researchers improved complexity bounds from PSPACE to CH, the counting hierarchy.

This improvement applies to deciding if a Chow form vanishes at a point. It also applies to checking if a monomial is in the support of a Chow form.

The bounds apply to membership in an equidimensional component. They also apply to membership in the closure of the image of an irreducible variety under a polynomial map.

The researchers constructed a polylogtime-uniform family of constant-depth, exponential-size threshold circuits. These circuits compute the Chow forms of the equidimensional components.

They built these circuits by combining geometric resolutions with the machinery of Andrews, Garg, and Schost (FOCS 2026) for the multivariate resultant.

Key facts

  • Complexity bounds for equidimensional decomposition were improved from PSPACE to CH.
  • The approximate polynomial satisfiability problem can be decided in CH.
  • The degree of the kthk^{\text{th}} equidimensional component of a projective variety can be computed in FPCH\text{FP}^{\text{CH}}.

All sections

Edition