QMBLthe quantum many-body leaderboard

All instances / Hubbard

Hubbard square 4x4, open, U = 8, n = 1

QMBL lists 21 published ground-state energies for the Hubbard model (square 4x4, open, U = 8, n = 1). The exact energy is -0.42552590415, by ED in QMBL check, checks/hubbard-4x4-open-ed/; and the best variational energy is -0.4255259, by DMRG in Liu et al. (2025). Energies as E/site.

Exact energy and best values (rules §6)

exact -0.42552590415 ED
QMBL check, checks/hubbard-4x4-open-ed/ · the instance is solved
best variational -0.4255259 DMRG
Liu et al. (2025) · the lowest unflagged variational energy, a strict variational bound

Energies as E/site. Classes are different claims and are not ranked against each other.

The Hamiltonian

H = −t ∑⟨ij⟩,σ (c†iσcjσ + h.c.) + U ∑i ni↑ni↓

t = 1, U = 8, N↑ = N↓ = 8 on 16 sites.

lattice
square
sites
16
boundary
open
Nf
8
U
8
energies
21

Stored as a total energy; quoted here as E/site, which is what the papers report (how the conversion works).

Every published energy

Grouped by what the number is. The exact energy comes first where there is one; in every other group the lowest unflagged energy is its best value, and each row's distance is to that best (rules §6).

Numerically exact (1)

Exact diagonalization, an exact solution, or sign-problem-free QMC where that is established. An exact energy is shown first, before the best of each other class. QMC is exact only within its error bar, which it must state, and reads exact (stochastic); an exact diagonalization outranks it. For a diagonalization resolved by symmetry sector only the ground state's sector is listed; the lowest energies of the other sectors are the instance's spectrum, in its JSON, and are not shown as the exact energy.

E/sitedistance to class bestσVar(E)V-score methodsourcecodeyearcomment
-0.42552590415exact 0 n/a n/a n/a ED
QMBL run
QMBL check, checks/hubbard-4x4-open-ed/ QMBL code 2026

Strict variational upper bounds (19)

Energies that a variational state reached (VMC, DMRG and MPS at finite bond dimension, PEPS, NQS, VQE). Each is an upper bound on the ground-state energy.

E/sitedistance to class bestσVar(E)V-score methodsourcecodeyearcomment
-0.4255259best variational 0 n/a n/a n/a DMRG
peer reviewed
Liu et al. (2025) hczhai/2d-hubbard-dmrg-2024 2025
-0.4255259 <1e-7 n/a n/a n/a DMRG
peer reviewed
Liu et al. (2025) hczhai/2d-hubbard-dmrg-2024 2025
-0.4255258 <1e-7 n/a n/a n/a DMRG
peer reviewed
Liu et al. (2025) hczhai/2d-hubbard-dmrg-2024 2025
-0.4255252 7.1e-7 n/a n/a n/a DMRG
peer reviewed
Liu et al. (2025) hczhai/2d-hubbard-dmrg-2024 2025
-0.425519tied with best variational at 2σ 6.9e-6 6.0e-6 n/a n/a VMC (fPEPS)
peer reviewed
Liu et al. (2025) 2025
-0.425515 1.1e-5 n/a n/a n/a DMRG
peer reviewed
Liu et al. (2025) hczhai/2d-hubbard-dmrg-2024 2025
-0.425508 1.8e-5 5.0e-6 n/a n/a VMC (fPEPS)
peer reviewed
Liu et al. (2025) 2025
-0.425493 3.3e-5 3.0e-6 n/a n/a VMC (fPEPS)
peer reviewed
Liu et al. (2025) 2025
-0.425429 9.7e-5 8.0e-6 n/a n/a VMC (fPEPS)
peer reviewed
Liu et al. (2025) 2025
-0.42508 4.5e-4 4.0e-5 n/a n/a VMC (fPEPS)
peer reviewed
Liu et al. (2025) 2025
-0.4244 1.1e-3 2.0e-4 n/a n/a simple update (fPEPS)
peer reviewed
Liu et al. (2025) 2025
-0.4242 1.3e-3 5.0e-4 n/a n/a simple update (fPEPS)
peer reviewed
Liu et al. (2025) 2025
-0.42416 1.4e-3 7.0e-5 n/a n/a VMC (fPEPS)
peer reviewed
Liu et al. (2025) 2025
-0.4241 1.4e-3 3.0e-4 n/a n/a simple update (fPEPS)
peer reviewed
Liu et al. (2025) 2025
-0.4238 1.7e-3 5.0e-4 n/a n/a simple update (fPEPS)
peer reviewed
Liu et al. (2025) 2025
-0.4236 1.9e-3 3.0e-4 n/a n/a simple update (fPEPS)
peer reviewed
Liu et al. (2025) 2025
-0.42115 4.4e-3 3.0e-5 n/a n/a VMC (fPEPS)
peer reviewed
Liu et al. (2025) 2025
-0.4204 5.1e-3 6.0e-4 n/a n/a simple update (fPEPS)
peer reviewed
Liu et al. (2025) 2025
-0.4174 8.1e-3 2.0e-4 n/a n/a simple update (fPEPS)
peer reviewed
Liu et al. (2025) 2025

Extrapolations (1)

Zero-variance or bond-dimension extrapolations. These are not bounds: no ansatz reached the number.

E/sitedistance to class bestσVar(E)V-score methodsourcecodeyearcomment
-0.42552596 n/a 1.0e-8 n/a n/a DMRG
below exact by > 2σ peer reviewed
Liu et al. (2025) hczhai/2d-hubbard-dmrg-2024 2025

How these numbers were read

Where each number below was read, and how it was checked.

-0.42552596(1), DMRG

Checked 2026-11-09 · source text of arXiv:2502.13454 read locally (arXiv HTML or pypdf layout text); value copied from the harvested cell and the text, no LLM transcription of numbers; single reading

Reported as. -0.42552596(1)

SM Table I, 4 × 4, half filling (Ne = 16), row D = 14 / m = ∞; dl/liu-sm.txt line 692 (journal Supplemental Material), column DMRG | H = -t sum_<ij>,s (c+ c + h.c.) + U sum n_up n_dn (PRL Eq. 1), t = 1, U = 8; open boundary conditions along both x and y; energies per site. fPEPS: simple update (SU) and gradient optimisation (GO) with Monte Carlo sampling, error = Monte Carlo sampling error; DMRG: SU(2) DMRG (block2) at m multiplets, m = infinity extrapolated in the truncated (discarded) weight.

-0.42552590415, ED

Checked 2026-11-12 · ed_full2.py --obc (code B: plain Lanczos, Hubbard kernel with fixed N_up = N_dn = 8, 165,636,900 states, true residual 7.0e-11) and hub_fact.py --obc (code A: the factorised matvec T x 1 + 1 x T + U D of checks/hubbard-u-labels/ed_check.py with ARPACK, residual 4.1e-12), an AMD EPYC 9654 server, 2026-11-12

Reported as. -6.80841446644252 (ed_full2.py) | -6.808414466441927 (hub_fact.py)

Total energy, t = 1, H = -t sum (c+c + h.c.) + U sum n_up n_dn on the open 4x4 lattice (24 bonds), N_up = N_dn = 8 (half filling; Lieb's theorem puts the unique singlet in this sector); E/N = -0.425525904. No published exact energy is carried for this instance; the two codes agree to 5.9e-13. Liu et al.'s DMRG extrapolated in the truncated weight, -6.80841536 (sigma 1.6e-7), lies 8.9e-7 below, 5.6 sigma.

-0.4255259, DMRG

Checked 2026-11-09 · source text of arXiv:2502.13454 read locally (arXiv HTML or pypdf layout text); value copied from the harvested cell and the text, no LLM transcription of numbers; single reading

Reported as. -0.42552590

SM Table I, 4 × 4, half filling (Ne = 16), row D = 12 / m = 2500; dl/liu-sm.txt line 691 (journal Supplemental Material), column DMRG | H = -t sum_<ij>,s (c+ c + h.c.) + U sum n_up n_dn (PRL Eq. 1), t = 1, U = 8; open boundary conditions along both x and y; energies per site. fPEPS: simple update (SU) and gradient optimisation (GO) with Monte Carlo sampling, error = Monte Carlo sampling error; DMRG: SU(2) DMRG (block2) at m multiplets, m = infinity extrapolated in the truncated (discarded) weight.

-0.4255259, DMRG

Checked 2026-11-09 · source text of arXiv:2502.13454 read locally (arXiv HTML or pypdf layout text); value copied from the harvested cell and the text, no LLM transcription of numbers; single reading

Reported as. -0.42552589

SM Table I, 4 × 4, half filling (Ne = 16), row D = 10 / m = 2000; dl/liu-sm.txt line 690 (journal Supplemental Material), column DMRG | H = -t sum_<ij>,s (c+ c + h.c.) + U sum n_up n_dn (PRL Eq. 1), t = 1, U = 8; open boundary conditions along both x and y; energies per site. fPEPS: simple update (SU) and gradient optimisation (GO) with Monte Carlo sampling, error = Monte Carlo sampling error; DMRG: SU(2) DMRG (block2) at m multiplets, m = infinity extrapolated in the truncated (discarded) weight.

-0.4255258, DMRG

Checked 2026-11-09 · source text of arXiv:2502.13454 read locally (arXiv HTML or pypdf layout text); value copied from the harvested cell and the text, no LLM transcription of numbers; single reading

Reported as. -0.42552583

SM Table I, 4 × 4, half filling (Ne = 16), row D = 8 / m = 1500; dl/liu-sm.txt line 689 (journal Supplemental Material), column DMRG | H = -t sum_<ij>,s (c+ c + h.c.) + U sum n_up n_dn (PRL Eq. 1), t = 1, U = 8; open boundary conditions along both x and y; energies per site. fPEPS: simple update (SU) and gradient optimisation (GO) with Monte Carlo sampling, error = Monte Carlo sampling error; DMRG: SU(2) DMRG (block2) at m multiplets, m = infinity extrapolated in the truncated (discarded) weight.

-0.4255252, DMRG

Checked 2026-11-09 · source text of arXiv:2502.13454 read locally (arXiv HTML or pypdf layout text); value copied from the harvested cell and the text, no LLM transcription of numbers; single reading

Reported as. -0.42552519

SM Table I, 4 × 4, half filling (Ne = 16), row D = 6 / m = 1000; dl/liu-sm.txt line 688 (journal Supplemental Material), column DMRG | H = -t sum_<ij>,s (c+ c + h.c.) + U sum n_up n_dn (PRL Eq. 1), t = 1, U = 8; open boundary conditions along both x and y; energies per site. fPEPS: simple update (SU) and gradient optimisation (GO) with Monte Carlo sampling, error = Monte Carlo sampling error; DMRG: SU(2) DMRG (block2) at m multiplets, m = infinity extrapolated in the truncated (discarded) weight.

-0.425519(6), VMC (fPEPS)

Checked 2026-11-09 · source text of arXiv:2502.13454 read locally (arXiv HTML or pypdf layout text); value copied from the harvested cell and the text, no LLM transcription of numbers; single reading

Reported as. -0.425519(6)

SM Table I, 4 × 4, half filling (Ne = 16), row D = 16; dl/liu-sm.txt line 693 (journal Supplemental Material), column fPEPS-GO | H = -t sum_<ij>,s (c+ c + h.c.) + U sum n_up n_dn (PRL Eq. 1), t = 1, U = 8; open boundary conditions along both x and y; energies per site. fPEPS: simple update (SU) and gradient optimisation (GO) with Monte Carlo sampling, error = Monte Carlo sampling error; DMRG: SU(2) DMRG (block2) at m multiplets, m = infinity extrapolated in the truncated (discarded) weight.

-0.425515, DMRG

Checked 2026-11-09 · source text of arXiv:2502.13454 read locally (arXiv HTML or pypdf layout text); value copied from the harvested cell and the text, no LLM transcription of numbers; single reading

Reported as. -0.42551504

SM Table I, 4 × 4, half filling (Ne = 16), row D = 4 / m = 500; dl/liu-sm.txt line 687 (journal Supplemental Material), column DMRG | H = -t sum_<ij>,s (c+ c + h.c.) + U sum n_up n_dn (PRL Eq. 1), t = 1, U = 8; open boundary conditions along both x and y; energies per site. fPEPS: simple update (SU) and gradient optimisation (GO) with Monte Carlo sampling, error = Monte Carlo sampling error; DMRG: SU(2) DMRG (block2) at m multiplets, m = infinity extrapolated in the truncated (discarded) weight.

-0.425508(5), VMC (fPEPS)

Checked 2026-11-09 · source text of arXiv:2502.13454 read locally (arXiv HTML or pypdf layout text); value copied from the harvested cell and the text, no LLM transcription of numbers; single reading

Reported as. -0.425508(5)

SM Table I, 4 × 4, half filling (Ne = 16), row D = 14 / m = ∞; dl/liu-sm.txt line 692 (journal Supplemental Material), column fPEPS-GO | H = -t sum_<ij>,s (c+ c + h.c.) + U sum n_up n_dn (PRL Eq. 1), t = 1, U = 8; open boundary conditions along both x and y; energies per site. fPEPS: simple update (SU) and gradient optimisation (GO) with Monte Carlo sampling, error = Monte Carlo sampling error; DMRG: SU(2) DMRG (block2) at m multiplets, m = infinity extrapolated in the truncated (discarded) weight.

-0.425493(3), VMC (fPEPS)

Checked 2026-11-09 · source text of arXiv:2502.13454 read locally (arXiv HTML or pypdf layout text); value copied from the harvested cell and the text, no LLM transcription of numbers; single reading

Reported as. -0.425493(3)

SM Table I, 4 × 4, half filling (Ne = 16), row D = 12 / m = 2500; dl/liu-sm.txt line 691 (journal Supplemental Material), column fPEPS-GO | H = -t sum_<ij>,s (c+ c + h.c.) + U sum n_up n_dn (PRL Eq. 1), t = 1, U = 8; open boundary conditions along both x and y; energies per site. fPEPS: simple update (SU) and gradient optimisation (GO) with Monte Carlo sampling, error = Monte Carlo sampling error; DMRG: SU(2) DMRG (block2) at m multiplets, m = infinity extrapolated in the truncated (discarded) weight.

-0.425429(8), VMC (fPEPS)

Checked 2026-11-09 · source text of arXiv:2502.13454 read locally (arXiv HTML or pypdf layout text); value copied from the harvested cell and the text, no LLM transcription of numbers; single reading

Reported as. -0.425429(8)

SM Table I, 4 × 4, half filling (Ne = 16), row D = 10 / m = 2000; dl/liu-sm.txt line 690 (journal Supplemental Material), column fPEPS-GO | H = -t sum_<ij>,s (c+ c + h.c.) + U sum n_up n_dn (PRL Eq. 1), t = 1, U = 8; open boundary conditions along both x and y; energies per site. fPEPS: simple update (SU) and gradient optimisation (GO) with Monte Carlo sampling, error = Monte Carlo sampling error; DMRG: SU(2) DMRG (block2) at m multiplets, m = infinity extrapolated in the truncated (discarded) weight.

-0.42508(4), VMC (fPEPS)

Checked 2026-11-09 · source text of arXiv:2502.13454 read locally (arXiv HTML or pypdf layout text); value copied from the harvested cell and the text, no LLM transcription of numbers; single reading

Reported as. -0.42508(4)

SM Table I, 4 × 4, half filling (Ne = 16), row D = 8 / m = 1500; dl/liu-sm.txt line 689 (journal Supplemental Material), column fPEPS-GO | H = -t sum_<ij>,s (c+ c + h.c.) + U sum n_up n_dn (PRL Eq. 1), t = 1, U = 8; open boundary conditions along both x and y; energies per site. fPEPS: simple update (SU) and gradient optimisation (GO) with Monte Carlo sampling, error = Monte Carlo sampling error; DMRG: SU(2) DMRG (block2) at m multiplets, m = infinity extrapolated in the truncated (discarded) weight.

-0.4244(2), simple update (fPEPS)

Checked 2026-11-09 · source text of arXiv:2502.13454 read locally (arXiv HTML or pypdf layout text); value copied from the harvested cell and the text, no LLM transcription of numbers; single reading

Reported as. -0.4244(2)

SM Table I, 4 × 4, half filling (Ne = 16), row D = 16; dl/liu-sm.txt line 693 (journal Supplemental Material), column fPEPS-SU | H = -t sum_<ij>,s (c+ c + h.c.) + U sum n_up n_dn (PRL Eq. 1), t = 1, U = 8; open boundary conditions along both x and y; energies per site. fPEPS: simple update (SU) and gradient optimisation (GO) with Monte Carlo sampling, error = Monte Carlo sampling error; DMRG: SU(2) DMRG (block2) at m multiplets, m = infinity extrapolated in the truncated (discarded) weight.

-0.4242(5), simple update (fPEPS)

Checked 2026-11-09 · source text of arXiv:2502.13454 read locally (arXiv HTML or pypdf layout text); value copied from the harvested cell and the text, no LLM transcription of numbers; single reading

Reported as. -0.4242(5)

SM Table I, 4 × 4, half filling (Ne = 16), row D = 14 / m = ∞; dl/liu-sm.txt line 692 (journal Supplemental Material), column fPEPS-SU | H = -t sum_<ij>,s (c+ c + h.c.) + U sum n_up n_dn (PRL Eq. 1), t = 1, U = 8; open boundary conditions along both x and y; energies per site. fPEPS: simple update (SU) and gradient optimisation (GO) with Monte Carlo sampling, error = Monte Carlo sampling error; DMRG: SU(2) DMRG (block2) at m multiplets, m = infinity extrapolated in the truncated (discarded) weight.

-0.42416(7), VMC (fPEPS)

Checked 2026-11-09 · source text of arXiv:2502.13454 read locally (arXiv HTML or pypdf layout text); value copied from the harvested cell and the text, no LLM transcription of numbers; single reading

Reported as. -0.42416(7)

SM Table I, 4 × 4, half filling (Ne = 16), row D = 6 / m = 1000; dl/liu-sm.txt line 688 (journal Supplemental Material), column fPEPS-GO | H = -t sum_<ij>,s (c+ c + h.c.) + U sum n_up n_dn (PRL Eq. 1), t = 1, U = 8; open boundary conditions along both x and y; energies per site. fPEPS: simple update (SU) and gradient optimisation (GO) with Monte Carlo sampling, error = Monte Carlo sampling error; DMRG: SU(2) DMRG (block2) at m multiplets, m = infinity extrapolated in the truncated (discarded) weight.

-0.4241(3), simple update (fPEPS)

Checked 2026-11-09 · source text of arXiv:2502.13454 read locally (arXiv HTML or pypdf layout text); value copied from the harvested cell and the text, no LLM transcription of numbers; single reading

Reported as. -0.4241(3)

SM Table I, 4 × 4, half filling (Ne = 16), row D = 12 / m = 2500; dl/liu-sm.txt line 691 (journal Supplemental Material), column fPEPS-SU | H = -t sum_<ij>,s (c+ c + h.c.) + U sum n_up n_dn (PRL Eq. 1), t = 1, U = 8; open boundary conditions along both x and y; energies per site. fPEPS: simple update (SU) and gradient optimisation (GO) with Monte Carlo sampling, error = Monte Carlo sampling error; DMRG: SU(2) DMRG (block2) at m multiplets, m = infinity extrapolated in the truncated (discarded) weight.

-0.4238(5), simple update (fPEPS)

Checked 2026-11-09 · source text of arXiv:2502.13454 read locally (arXiv HTML or pypdf layout text); value copied from the harvested cell and the text, no LLM transcription of numbers; single reading

Reported as. -0.4238(5)

SM Table I, 4 × 4, half filling (Ne = 16), row D = 8 / m = 1500; dl/liu-sm.txt line 689 (journal Supplemental Material), column fPEPS-SU | H = -t sum_<ij>,s (c+ c + h.c.) + U sum n_up n_dn (PRL Eq. 1), t = 1, U = 8; open boundary conditions along both x and y; energies per site. fPEPS: simple update (SU) and gradient optimisation (GO) with Monte Carlo sampling, error = Monte Carlo sampling error; DMRG: SU(2) DMRG (block2) at m multiplets, m = infinity extrapolated in the truncated (discarded) weight.

-0.4236(3), simple update (fPEPS)

Checked 2026-11-09 · source text of arXiv:2502.13454 read locally (arXiv HTML or pypdf layout text); value copied from the harvested cell and the text, no LLM transcription of numbers; single reading

Reported as. -0.4236(3)

SM Table I, 4 × 4, half filling (Ne = 16), row D = 10 / m = 2000; dl/liu-sm.txt line 690 (journal Supplemental Material), column fPEPS-SU | H = -t sum_<ij>,s (c+ c + h.c.) + U sum n_up n_dn (PRL Eq. 1), t = 1, U = 8; open boundary conditions along both x and y; energies per site. fPEPS: simple update (SU) and gradient optimisation (GO) with Monte Carlo sampling, error = Monte Carlo sampling error; DMRG: SU(2) DMRG (block2) at m multiplets, m = infinity extrapolated in the truncated (discarded) weight.

-0.42115(3), VMC (fPEPS)

Checked 2026-11-09 · source text of arXiv:2502.13454 read locally (arXiv HTML or pypdf layout text); value copied from the harvested cell and the text, no LLM transcription of numbers; single reading

Reported as. -0.42115(3)

SM Table I, 4 × 4, half filling (Ne = 16), row D = 4 / m = 500; dl/liu-sm.txt line 687 (journal Supplemental Material), column fPEPS-GO | H = -t sum_<ij>,s (c+ c + h.c.) + U sum n_up n_dn (PRL Eq. 1), t = 1, U = 8; open boundary conditions along both x and y; energies per site. fPEPS: simple update (SU) and gradient optimisation (GO) with Monte Carlo sampling, error = Monte Carlo sampling error; DMRG: SU(2) DMRG (block2) at m multiplets, m = infinity extrapolated in the truncated (discarded) weight.

-0.4204(6), simple update (fPEPS)

Checked 2026-11-09 · source text of arXiv:2502.13454 read locally (arXiv HTML or pypdf layout text); value copied from the harvested cell and the text, no LLM transcription of numbers; single reading

Reported as. -0.4204(6)

SM Table I, 4 × 4, half filling (Ne = 16), row D = 6 / m = 1000; dl/liu-sm.txt line 688 (journal Supplemental Material), column fPEPS-SU | H = -t sum_<ij>,s (c+ c + h.c.) + U sum n_up n_dn (PRL Eq. 1), t = 1, U = 8; open boundary conditions along both x and y; energies per site. fPEPS: simple update (SU) and gradient optimisation (GO) with Monte Carlo sampling, error = Monte Carlo sampling error; DMRG: SU(2) DMRG (block2) at m multiplets, m = infinity extrapolated in the truncated (discarded) weight.

-0.4174(2), simple update (fPEPS)

Checked 2026-11-09 · source text of arXiv:2502.13454 read locally (arXiv HTML or pypdf layout text); value copied from the harvested cell and the text, no LLM transcription of numbers; single reading

Reported as. -0.4174(2)

SM Table I, 4 × 4, half filling (Ne = 16), row D = 4 / m = 500; dl/liu-sm.txt line 687 (journal Supplemental Material), column fPEPS-SU | H = -t sum_<ij>,s (c+ c + h.c.) + U sum n_up n_dn (PRL Eq. 1), t = 1, U = 8; open boundary conditions along both x and y; energies per site. fPEPS: simple update (SU) and gradient optimisation (GO) with Monte Carlo sampling, error = Monte Carlo sampling error; DMRG: SU(2) DMRG (block2) at m multiplets, m = infinity extrapolated in the truncated (discarded) weight.

Coverage

Newest published number here: 2025.

This instance as data

JSON · source file on GitHub · add a result · open an issue about this instance