QMBLthe quantum many-body leaderboard

All instances / Hydrogen chain

Hydrogen chain H50 chain, open, R = 2.8, STO-6G

Best values (rules §6)

best variational -0.505496 LDMRG
10.1063/1.2345196 · the lowest unflagged variational energy, a strict variational bound
best projected -0.5056322 DMET
10.1103/PhysRevX.7.031059 · the lowest unflagged projected energy, a projected estimate, which depends on its node or constraint
best approximate -0.503953 UCCSD(T)
10.1103/PhysRevX.7.031059 · the lowest unflagged approximate energy, an approximate method, whose error has no known sign

Energies as E/N (Ha). Classes are different claims and are not ranked against each other.

The Hamiltonian

H = −½ ∑i ∇i2 − ∑i,I 1/|ri − RI| + ∑i<j 1/|ri − rj| + ∑I<J 1/|RI − RJ|

Hydrogen chain: 50 protons equally spaced on a line, spacing R = 2.8 bohr, open (a finite molecule); N↑ = 25, N↓ = 25. Projected onto the STO-6G basis set; full CI in it is this instance's exact energy. Hartree atomic units, nuclear repulsion included; the energy per atom is E/N.

geometry
chain
atoms
50
boundary
open
N_up
25
N_dn
25
R
2.8
basis
STO-6G
energies
14

Stored as a total energy in Hartree, nuclear repulsion included (rules §13); quoted here as E/N (Ha), the energy per atom (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).

Strict variational upper bounds (6)

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/N (Ha)distance to class bestσVar(E)V-score methodsourcecodeyearcomment
-0.505496best variational 0 n/a n/a n/a LDMRG
peer reviewed
 10.1063/1.2345196 2006
-0.5054958 <1e-6 n/a n/a n/a DMRG
peer reviewed
 10.1103/PhysRevX.7.031059 2017
-0.5054958 <1e-6 n/a n/a n/a LDMRG
peer reviewed
 10.1063/1.2345196 2006
-0.5054924 3.6e-6 n/a n/a n/a LDMRG
peer reviewed
 10.1063/1.2345196 2006
-0.4926837 1.3e-2 n/a n/a n/a HF
peer reviewed
 10.1103/PhysRevX.7.031059 2017
-0.4672162 3.8e-2 n/a n/a n/a HF
peer reviewed
 10.1063/1.2345196 2006

Projected or fixed-node estimates (5)

Variational only within a constraint: fixed-node, constrained-path, GFMC on a trial state. Each depends on its node or constraint, so two of them are not cleanly comparable to each other or to strict bounds.

E/N (Ha)distance to class bestσVar(E)V-score methodsourcecodeyearcomment
-0.5056322best projected 0 n/a n/a n/a DMET
peer reviewed
 10.1103/PhysRevX.7.031059 2017
-0.50553 1.0e-4 1.4e-5 n/a n/a ph-AFQMC
peer reviewed
 10.1103/PhysRevX.7.031059 2017
-0.5048644 7.7e-4 n/a n/a n/a SEET
peer reviewed
 10.1103/PhysRevX.7.031059 2017
-0.5038132 1.8e-3 n/a n/a n/a UCCSD
peer reviewed
 10.1103/PhysRevX.7.031059 2017
-0.5026845 2.9e-3 n/a n/a n/a DMET
peer reviewed
 10.1103/PhysRevX.7.031059 2017

Approximate methods (no bound) (3)

Methods that converge to a fixed approximation and not to the exact energy: RPA, finite-order perturbation theory, coupled cluster restricted to some of its terms. Their error has no known sign, so they are compared only with each other (rules §4).

E/N (Ha)distance to class bestσVar(E)V-score methodsourcecodeyearcomment
-0.503953best approximate 0 n/a n/a n/a UCCSD(T)
peer reviewed
 10.1103/PhysRevX.7.031059 2017
-0.489934 1.4e-2 n/a n/a n/a GW
peer reviewed
 10.1103/PhysRevX.7.031059 2017
-0.4871222 1.7e-2 n/a n/a n/a MP2
peer reviewed
 10.1063/1.2345196 2006

How these numbers were read

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

-0.5056322, DMET

Checked 2026-10-20 · dataset github.com/simonsfoundation/hydrogen-benchmark-PRX, commit b99a183 (2017-10-09, uploaded by the authors), CC BY 4.0; source text of arXiv:1705.01608 read locally (pypdf layout text; the generator finds every value again on its line), no LLM transcription of numbers; single reading

Reported as. -25.281610 (total for N = 50)

dataset github.com/simonsfoundation/hydrogen-benchmark-PRX, commit b99a183 (2017-10-09, uploaded by the authors), CC BY 4.0, file N_50_OBC/R_2.8/DMET_5_basis-STO in sources/1705.01608-data-b99a183.tsv; number first printed in the dataset (RULES.md §8), the paper is the method reference | Hartree, lengths in bohr (Appendix B). Total energies with nuclear repulsion: FCI/STO-6G H10 at R = 1.8 is -5.42438538 in the dataset and -0.542439 per atom in Table II, and the paper's FCI E_inf = -0.471039 is the STO-6G hydrogen-atom energy per atom.

-0.50553(1), ph-AFQMC

Checked 2026-10-20 · dataset github.com/simonsfoundation/hydrogen-benchmark-PRX, commit b99a183 (2017-10-09, uploaded by the authors), CC BY 4.0; source text of arXiv:1705.01608 read locally (pypdf layout text; the generator finds every value again on its line), no LLM transcription of numbers; single reading

Reported as. -25.2765 0.0007 (total for N = 50; second number the stated uncertainty)

dataset github.com/simonsfoundation/hydrogen-benchmark-PRX, commit b99a183 (2017-10-09, uploaded by the authors), CC BY 4.0, file N_50_OBC/R_2.8/AFQMC_basis-STO in sources/1705.01608-data-b99a183.tsv; number first printed in the dataset (RULES.md §8), the paper is the method reference | Hartree, lengths in bohr (Appendix B). Total energies with nuclear repulsion: FCI/STO-6G H10 at R = 1.8 is -5.42438538 in the dataset and -0.542439 per atom in Table II, and the paper's FCI E_inf = -0.471039 is the STO-6G hydrogen-atom energy per atom.

-0.505496, LDMRG

Checked 2026-10-20 · source text of arXiv:cond-mat/0606115 read locally (pypdf layout text; the generator finds every value again on its line), no LLM transcription of numbers; single reading

Reported as. E_RHF -23.36081 + E_corr -1.91399 = -25.27480 (totals)

Table VI, R = 2.8 row; sources/cond-mat-0606115.txt line 835; version read: arXiv v3 (journal copy not reached, AIP HTTP 403) | Hartree totals: E_RHF (total, with nuclear repulsion) and E_RHF,el printed side by side; at R = 1.8 bohr E_RHF + E_corr[LDMRG(500)] = -27.03865 against the 2017 STO-6G DMRG -27.03864 for H50.

-0.5054958, DMRG

Checked 2026-10-20 · dataset github.com/simonsfoundation/hydrogen-benchmark-PRX, commit b99a183 (2017-10-09, uploaded by the authors), CC BY 4.0; source text of arXiv:1705.01608 read locally (pypdf layout text; the generator finds every value again on its line), no LLM transcription of numbers; single reading

Reported as. -25.27479 (total for N = 50)

dataset github.com/simonsfoundation/hydrogen-benchmark-PRX, commit b99a183 (2017-10-09, uploaded by the authors), CC BY 4.0, file N_50_OBC/R_2.8/DMRG_basis-STO in sources/1705.01608-data-b99a183.tsv; number first printed in the dataset (RULES.md §8), the paper is the method reference | Hartree, lengths in bohr (Appendix B). Total energies with nuclear repulsion: FCI/STO-6G H10 at R = 1.8 is -5.42438538 in the dataset and -0.542439 per atom in Table II, and the paper's FCI E_inf = -0.471039 is the STO-6G hydrogen-atom energy per atom. | Read as the variational energy of the final sweep at bond dimension D = 2000 (Appendix A runs Block with the standard settings of Ref. [66], whose schedule ends at the maximum M; the paper names no extrapolation in the bond dimension). Medium confidence: Ref. [66] also describes extrapolation to zero discarded weight, and the paper does not say whether these energies were extrapolated (ruled 2026-10-24). | reader confidence medium

-0.5054958, LDMRG

Checked 2026-10-20 · source text of arXiv:cond-mat/0606115 read locally (pypdf layout text; the generator finds every value again on its line), no LLM transcription of numbers; single reading

Reported as. E_RHF -23.36081 + E_corr -1.91398 = -25.27479 (totals)

Table VI, R = 2.8 row; sources/cond-mat-0606115.txt line 835; version read: arXiv v3 (journal copy not reached, AIP HTTP 403) | Hartree totals: E_RHF (total, with nuclear repulsion) and E_RHF,el printed side by side; at R = 1.8 bohr E_RHF + E_corr[LDMRG(500)] = -27.03865 against the 2017 STO-6G DMRG -27.03864 for H50.

-0.5054924, LDMRG

Checked 2026-10-20 · source text of arXiv:cond-mat/0606115 read locally (pypdf layout text; the generator finds every value again on its line), no LLM transcription of numbers; single reading

Reported as. E_RHF -23.36081 + E_corr -1.91381 = -25.27462 (totals)

Table VI, R = 2.8 row; sources/cond-mat-0606115.txt line 835; version read: arXiv v3 (journal copy not reached, AIP HTTP 403) | Hartree totals: E_RHF (total, with nuclear repulsion) and E_RHF,el printed side by side; at R = 1.8 bohr E_RHF + E_corr[LDMRG(500)] = -27.03865 against the 2017 STO-6G DMRG -27.03864 for H50.

-0.5048644, SEET

Checked 2026-10-20 · dataset github.com/simonsfoundation/hydrogen-benchmark-PRX, commit b99a183 (2017-10-09, uploaded by the authors), CC BY 4.0; source text of arXiv:1705.01608 read locally (pypdf layout text; the generator finds every value again on its line), no LLM transcription of numbers; single reading

Reported as. -25.24321959 (total for N = 50)

dataset github.com/simonsfoundation/hydrogen-benchmark-PRX, commit b99a183 (2017-10-09, uploaded by the authors), CC BY 4.0, file N_50_OBC/R_2.8/SEET_CI-GF2_mix(6,SAO)_basis-STO in sources/1705.01608-data-b99a183.tsv; number first printed in the dataset (RULES.md §8), the paper is the method reference | Hartree, lengths in bohr (Appendix B). Total energies with nuclear repulsion: FCI/STO-6G H10 at R = 1.8 is -5.42438538 in the dataset and -0.542439 per atom in Table II, and the paper's FCI E_inf = -0.471039 is the STO-6G hydrogen-atom energy per atom.

-0.503953, UCCSD(T)

Checked 2026-10-20 · dataset github.com/simonsfoundation/hydrogen-benchmark-PRX, commit b99a183 (2017-10-09, uploaded by the authors), CC BY 4.0; source text of arXiv:1705.01608 read locally (pypdf layout text; the generator finds every value again on its line), no LLM transcription of numbers; single reading

Reported as. -25.197649 (total for N = 50)

dataset github.com/simonsfoundation/hydrogen-benchmark-PRX, commit b99a183 (2017-10-09, uploaded by the authors), CC BY 4.0, file N_50_OBC/R_2.8/UCCSD(T)_basis-STO in sources/1705.01608-data-b99a183.tsv; number first printed in the dataset (RULES.md §8), the paper is the method reference | Hartree, lengths in bohr (Appendix B). Total energies with nuclear repulsion: FCI/STO-6G H10 at R = 1.8 is -5.42438538 in the dataset and -0.542439 per atom in Table II, and the paper's FCI E_inf = -0.471039 is the STO-6G hydrogen-atom energy per atom.

-0.5038132, UCCSD

Checked 2026-10-20 · dataset github.com/simonsfoundation/hydrogen-benchmark-PRX, commit b99a183 (2017-10-09, uploaded by the authors), CC BY 4.0; source text of arXiv:1705.01608 read locally (pypdf layout text; the generator finds every value again on its line), no LLM transcription of numbers; single reading

Reported as. -25.190658 (total for N = 50)

dataset github.com/simonsfoundation/hydrogen-benchmark-PRX, commit b99a183 (2017-10-09, uploaded by the authors), CC BY 4.0, file N_50_OBC/R_2.8/UCCSD_basis-STO in sources/1705.01608-data-b99a183.tsv; number first printed in the dataset (RULES.md §8), the paper is the method reference | Hartree, lengths in bohr (Appendix B). Total energies with nuclear repulsion: FCI/STO-6G H10 at R = 1.8 is -5.42438538 in the dataset and -0.542439 per atom in Table II, and the paper's FCI E_inf = -0.471039 is the STO-6G hydrogen-atom energy per atom.

-0.5026845, DMET

Checked 2026-10-20 · dataset github.com/simonsfoundation/hydrogen-benchmark-PRX, commit b99a183 (2017-10-09, uploaded by the authors), CC BY 4.0; source text of arXiv:1705.01608 read locally (pypdf layout text; the generator finds every value again on its line), no LLM transcription of numbers; single reading

Reported as. -25.134224 (total for N = 50)

dataset github.com/simonsfoundation/hydrogen-benchmark-PRX, commit b99a183 (2017-10-09, uploaded by the authors), CC BY 4.0, file N_50_OBC/R_2.8/DMET_2_basis-STO in sources/1705.01608-data-b99a183.tsv; number first printed in the dataset (RULES.md §8), the paper is the method reference | Hartree, lengths in bohr (Appendix B). Total energies with nuclear repulsion: FCI/STO-6G H10 at R = 1.8 is -5.42438538 in the dataset and -0.542439 per atom in Table II, and the paper's FCI E_inf = -0.471039 is the STO-6G hydrogen-atom energy per atom.

-0.4926837, HF

Checked 2026-10-20 · dataset github.com/simonsfoundation/hydrogen-benchmark-PRX, commit b99a183 (2017-10-09, uploaded by the authors), CC BY 4.0; source text of arXiv:1705.01608 read locally (pypdf layout text; the generator finds every value again on its line), no LLM transcription of numbers; single reading

Reported as. -24.6341856684 (total for N = 50)

dataset github.com/simonsfoundation/hydrogen-benchmark-PRX, commit b99a183 (2017-10-09, uploaded by the authors), CC BY 4.0, file N_50_OBC/R_2.8/UHF_basis-STO in sources/1705.01608-data-b99a183.tsv; number first printed in the dataset (RULES.md §8), the paper is the method reference | Hartree, lengths in bohr (Appendix B). Total energies with nuclear repulsion: FCI/STO-6G H10 at R = 1.8 is -5.42438538 in the dataset and -0.542439 per atom in Table II, and the paper's FCI E_inf = -0.471039 is the STO-6G hydrogen-atom energy per atom.

-0.489934, GW

Checked 2026-10-20 · dataset github.com/simonsfoundation/hydrogen-benchmark-PRX, commit b99a183 (2017-10-09, uploaded by the authors), CC BY 4.0; source text of arXiv:1705.01608 read locally (pypdf layout text; the generator finds every value again on its line), no LLM transcription of numbers; single reading

Reported as. -24.4967 (total for N = 50)

dataset github.com/simonsfoundation/hydrogen-benchmark-PRX, commit b99a183 (2017-10-09, uploaded by the authors), CC BY 4.0, file N_50_OBC/R_2.8/SC-GW_basis-STO in sources/1705.01608-data-b99a183.tsv; number first printed in the dataset (RULES.md §8), the paper is the method reference | Hartree, lengths in bohr (Appendix B). Total energies with nuclear repulsion: FCI/STO-6G H10 at R = 1.8 is -5.42438538 in the dataset and -0.542439 per atom in Table II, and the paper's FCI E_inf = -0.471039 is the STO-6G hydrogen-atom energy per atom.

-0.4871222, MP2

Checked 2026-10-20 · source text of arXiv:cond-mat/0606115 read locally (pypdf layout text; the generator finds every value again on its line), no LLM transcription of numbers; single reading

Reported as. E_RHF -23.36081 + E_corr -0.99530 = -24.35611 (totals)

Table VI, R = 2.8 row; sources/cond-mat-0606115.txt line 835; version read: arXiv v3 (journal copy not reached, AIP HTTP 403) | Hartree totals: E_RHF (total, with nuclear repulsion) and E_RHF,el printed side by side; at R = 1.8 bohr E_RHF + E_corr[LDMRG(500)] = -27.03865 against the 2017 STO-6G DMRG -27.03864 for H50.

-0.4672162, HF

Checked 2026-10-20 · source text of arXiv:cond-mat/0606115 read locally (pypdf layout text; the generator finds every value again on its line), no LLM transcription of numbers; single reading

Reported as. E_RHF -23.360 81 (total)

Table VI, R = 2.8 row; sources/cond-mat-0606115.txt line 835; version read: arXiv v3 (journal copy not reached, AIP HTTP 403) | Hartree totals: E_RHF (total, with nuclear repulsion) and E_RHF,el printed side by side; at R = 1.8 bohr E_RHF + E_corr[LDMRG(500)] = -27.03865 against the 2017 STO-6G DMRG -27.03864 for H50.

Coverage

Newest published number here: 2017.

This instance as data

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