Polynomial-bond PEPS approximation of gapped grid ground states

Does every unique ground state of a gapped nearest-neighbor Hamiltonian on an open L×L square grid admit a projected entangled-pair state approximation of polynomial bond dimension? Local dimension q, interaction bound J and spectral gap Δ>0 are fixed; the target normalized vector error is L⁻¹.

Why it matters and what counts as progress

Why it matters: Polynomial bond dimension would provide a compact tensor-network representation for a broad class of two-dimensional gapped ground states.

What counts as progress: Prove bond dimension at most CLᶜ with constants controlled by q,J,Δ, or identify an obstruction under these same gap assumptions.

How the problem developed

  1. 29 Jan 2015 · Partial result [1]

    Norbert Schuch, Michael M. Wolf, Frank Verstraete and J. Ignacio Cirac Established efficient PEPS approximations for Gibbs states and ground states under a polynomial density-of-states assumption, not from a global gap alone.

  2. 24 Sept 2026 · Proof claim [2]

    OpenAI Released a manuscript claiming polynomial bond dimension and inverse-linear normalized-vector error for unique ground states of bounded local Hamiltonians on open square grids with a fixed global gap.

AI contribution

Polynomial PEPS approximation of gapped square-grid ground states

OpenAI

For fixed local dimension q, interaction bound J and global spectral gap Δ>0, the manuscript claims a PEPS with bond dimension at most CLᶜ and normalized vector error at most L⁻¹ for the unique ground state of any on-site plus nearest-neighbour Hamiltonian on an open L×L square grid.

AI involvementAI-led
ImportanceNotable
Technical verificationReported
Human understandingFew
Reported

Sources

  1. Schuch, Wolf, Verstraete and Cirac, Approximating Gibbs states of local Hamiltonians efficiently with projected entangled pair states
  2. OpenAI, polynomial PEPS approximation manuscript
  3. Polynomial PEPS approximation of gapped square-grid ground states
  4. OpenAI, companion two-dimensional area-law manuscript
  5. OpenAI Math release provenance
  6. Cirac and collaborators, related PEPS question
Website timeline

Publication dates and editorial changes.