{"id":"0012c117-2c90-4073-accf-e02ace82c86e","arxiv_id":"2506.05031","paper_version":2,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":4.0,"correctness_risk":"low","formal_verification":"none","parameter_count":3,"one_line_summary":"IQPE with a single Slater determinant reproduces exact ground-state energies for the Hubbard model on a six-site graphene hexagon in noiseless simulation, while hardware noise limits accuracy on current IBM devices.","lead":"Scientists used a quantum algorithm called IQPE to find the ground-state energy of a Hubbard model on a six-atom graphene hexagon, matching exact calculations in noiseless simulations. They then ran a smaller version on IBM quantum chips, where hardware noise degraded accuracy but one device still matched the exact answer.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The noiseless IQPE claim depends on the Slater determinant's ground-state overlap, which is never reported; without overlap values or measured success probabilities, the ED agreement could hide low-projection or excited-state outcomes.","rationale":"The reader's weakest assumption correctly identifies the missing overlap information as the key support gap. My read agrees: the noiseless central claim has independent support from the ED benchmarks and the convergence checks in Fig. 4, but the projection mechanism itself is treated as a black box. The overlap values are cheap to compute and would directly test whether the Slater determinant really has sufficient ground-state amplitude across all occupations and interaction strengths. The finite bit precision (m = 5) makes the ED agreement weaker than it appears because excited states lying within one phase bin would not be distinguished. The proposed concrete test settles this without requiring new physics. No internal inconsistency was found in the main numerical results, and the stated limitations (missing readout-robustness plot, ibm_fez error bars) are already captured by the reader's conditional verdict. Therefore the verdict remains CONDITIONAL and unchanged.","tokens_in":20403,"tokens_out":12557,"duration_ms":156567,"concrete_test":"Use exact diagonalization (QuSpin) to compute, for every Nocc and U0 used in Figs. 3 and 5, (i) the squared overlap between the prepared Slater determinant and the exact ground state, and (ii) the gap to the first excited state. Then rerun the noiseless IQPE circuit with a documented shot count (e.g., 10,000 shots per bit) and a fixed seed, and report the frequency with which the inferred energy bin matches the ED ground-state bin. If the minimum overlap is high (e.g., >0.5) and the ground-state bin is the majority outcome for all tested points, the concern is resolved. If overlap is low for some point, or if an excited-state bin wins, rerun with m = 6 and Ntrot = 30 to determine whether the reported agreement is an artifact of phase resolution or Trotter error.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim in Section III A (Figs. 3 and 5) is that a single Slater-determinant initial state allows IQPE to recover the exact ground-state energy for 1 <= Nocc <= 11 and U0 up to 6. IQPE is inherently projective: if the initial state is not an eigenstate, each run collapses onto an eigenstate with probability equal to the squared overlap. Section II B 1 explicitly states that a sufficiently large overlap is required, but the paper never reports the overlap values |<SD|GS>|^2 for any (Nocc, U0). The agreement with exact diagonalization is therefore indirect evidence. With m = 5 bits, the phase resolution is 2*pi/(32*t); when the gap to the first excited state is smaller than this resolution, an excited-state outcome can fall in the same energy bin as the ground state and still appear to match ED. The paper also does not specify whether the noiseless IQPE energy comes from a single run, majority voting over shots, or post-selection on the known ED answer; the last option would make the agreement circular. This is the weakest load-bearing point in the central claim, and it is directly checkable.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports quantum simulations of the Hubbard model on a six-site graphene hexagon using iterative quantum phase estimation (IQPE) and adiabatic evolution, implemented in Qiskit with the Jordan-Wigner mapping. In noiseless simulation, IQPE with a single Slater determinant initial state is shown to reproduce exact-diagonalization ground-state energies for occupation numbers 1 through 11 at U0=0 and U0=3, and for a range of interaction strengths up to U0=6 for a subset of occupations; convergence in bit precision m and Trotter steps N_trot is documented. The paper also presents a noise model based on ibm_strasbourg hardware parameters applied to a three-site system, isolating the effects of depolarizing errors, thermal relaxation, and readout errors, and reports hardware executions on ibm_strasbourg and ibm_fez. A simplification of the periodic-boundary Jordan-Wigner string to a global phase factor is derived and used to reduce circuit depth.","tokens_in":20674,"tokens_out":8475,"duration_ms":98615,"significance":"If the central claims hold, the paper provides a useful benchmark for IQPE on small Hubbard systems and a systematic noise analysis relevant to near-term hardware demonstrations. The noiseless agreement with exact diagonalization across many fillings and interaction strengths is a positive result, and the simplification of the Jordan-Wigner string is a concrete technical contribution. The convergence checks in m and N_trot add credibility. However, the main claim about the sufficiency of a single Slater determinant rests on a projective measurement protocol whose details and ground-state overlaps are not reported, and the ibm_fez hardware results are presented without statistical uncertainty. These gaps need to be addressed before the paper's conclusions are fully supported.","major_comments":[{"comment":"The central claim that a single Slater determinant is sufficient as an IQPE initial state is not fully supported because the manuscript never reports the squared overlaps |<SD|GS>|^2 for the studied (Nocc, U0) combinations, nor does it specify how the noiseless IQPE energy is extracted from the measured phase bits (e.g., a single run, repeated runs with majority voting, or post-selection). Since IQPE is projective, a low-overlap initial state can produce measurements of excited-state phases; with m=5 bits the phase resolution is 2*pi/32, so an excited state within the same phase bin could appear to match the ground-state energy. Reporting the overlaps and the measurement protocol is necessary to support the claim that the SD initialization works as stated.","section":"Section II B 1 and Section III A (Figs. 3-5)"},{"comment":"The ibm_fez hardware data are reported as an average over only 5 runs with 50,000 shots each and no error bars, whereas all other hardware and noisy-simulation results include standard deviations. The text states that the ibm_fez results 'closely match' the exact values; without a statistical uncertainty, the significance of this agreement cannot be assessed. The authors should report the standard deviation or variance for the ibm_fez data, or present the individual run values.","section":"Section III B, Fig. 8"}],"minor_comments":[{"comment":"The value of the time step t used in the IQPE simulations is never stated; it is only constrained by the phase-range condition. Please provide t for each simulation set so that the results are reproducible.","section":"Section III A"},{"comment":"The simplification of the Jordan-Wigner string in Eq. (10) relies on a fixed fermion number Nf in each spin sector; the manuscript should state explicitly that the Hubbard Hamiltonian conserves particle number per spin species, so that the phase factor is well defined within the chosen symmetry sector.","section":"Appendix, Eq. (A11)"},{"comment":"The legend entries in Fig. 4 are garbled ('ExactIQPE 4') and the axis labels in panel (a) appear partially cut off; please regenerate the figure so that all text is legible.","section":"Fig. 4"},{"comment":"In the description of the noise model, 'sing-qubit gate' should be 'single-qubit gate'.","section":"Section III B"},{"comment":"The manuscript states that the IQPE algorithm is robust to a wide range of readout errors but the supporting results are 'not shown here'; either include the corresponding figure or remove this claim from the text.","section":"Section III B"},{"comment":"For the adiabatic evolution results, the total evolution time T and the number of time-discretization steps used to implement the path H_ad(eta) are not reported; these parameters should be specified to support the claimed agreement with exact diagonalization.","section":"Section III A (adiabatic simulations)"}],"recommendation":"major_revision","confidential_remarks":"The overlap/protocol issue is the most important technical gap. If the authors can provide the squared overlaps and a clear description of how the IQPE energies were obtained (including any repetition or post-selection), the main noiseless-simulation claim can probably be verified. I would encourage the editor to request these additions along with the ibm_fez error bars. The paper is within the journal's scope and the topic is appropriate, but the missing statistical details currently prevent full support of the advertised hardware results."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nQuick take: this is a solid benchmark study of IQPE on a six-site Hubbard ring, and I think the central claim holds up. The noiseless results match exact diagonalization across all occupations 1–11 and U0 up to 6, with documented convergence in m and Ntrot. That is a genuine check, not a fit: nothing is tuned to the target energies. The JW-string-to-global-phase simplification for PBC is correct (the appendix proof works) and gives a real circuit-depth reduction. The noise study is careful, with variance over 50 runs and separate isolation of depolarizing, thermal, and readout channels.\n\nThe soft spots are real but not fatal. First, the paper never reports the squared overlap between the Slater determinant and the exact ground state, even though the algorithm's success depends on it. The stress-test worry is legitimate: IQPE is projective, and with m=5 bits an excited-state phase can land in the same energy bin if the gap is small. But the agreement across the full parameter sweep is strong circumstantial evidence that the overlaps are adequate; this is a missing check, not a demonstrated flaw. The authors should report overlaps and specify how the noiseless energies were aggregated (single run vs. majority vote vs. post-selection). Second, the text says readout errors are robust up to 0.2 but the data is 'not shown here'; that is an unsupported assertion and should be either shown or cut. Third, the ibm_fez hardware points have no error bars and average only 5 runs; that is a minor weakness given the hardware section is illustrative.\n\nThe citation pattern is fine; related work on IQPE and Slater determinants is properly credited. The paper is not groundbreaking—it is a benchmark on an eleven-qubit system—but it is a useful data point for people comparing IQPE to VQE on small Hubbard lattices, and the JW simplification is worth remembering.\n\nI would send this to peer review. The referee should ask for the overlap values and the readout plot, but the core results look solid.","headline":"Useful benchmark paper; central IQPE claim is credible, but the missing overlap data and unshown readout-robustness plot should be supplied before publication.","tokens_in":21175,"tokens_out":2689,"would_cite":false,"duration_ms":33875,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["81P68","82B20","81V70"],"pacs":["03.67.Lx","71.10.Fd"],"model":"deepseek-v4-flash","headline":"This paper shows that IQPE seeded by a single Slater determinant reproduces exact Hubbard ground-state energies on a six-site graphene hexagon, and identifies two-qubit gate noise and thermal relaxation as the main obstacles on real…","keywords":["Hubbard model","graphene hexagon","iterative quantum phase estimation","Slater determinant","Jordan-Wigner transformation","Trotter decomposition","noise model","ground-state energy"],"falsifier":"Compute the exact squared overlap between the prepare-and-evolve initial Slater determinant and the exact Hubbard ground state for a point inside the claimed range, say $N_{\\rm occ}=6$, $U_0=6$; if that overlap is essentially zero while noiseless IQPE still returns the exact ground-state energy with high probability on every run, the paper's stated mechanism for convergence would be falsified for that point.","tokens_in":20227,"feed_emoji":"⚛️","tokens_out":8425,"duration_ms":94024,"temperature":0.7,"pith_summary":"This paper sets out to show that Iterative Quantum Phase Estimation (IQPE), initialized with nothing more than a single Slater determinant, recovers the exact ground-state energy of the Hubbard model on a six-site graphene hexagon for the full tested range of occupation numbers and interaction strengths up to $U_0=6$. A Slater determinant is the cheap, non-interacting guess: a product of single-particle orbitals. The paper also argues that for this periodic six-site ring the boundary hopping term's Jordan-Wigner string collapses to a global phase factor that depends only on total particle number, which shortens the needed circuit. On the hardware side, the paper isolates which realistic error channels damage IQPE most, identifying two-qubit gate errors and thermal relaxation as the dominant obstacles and readout errors as nearly harmless. A run on a real superconducting device with improved parameters reproduces exact results for a reduced three-site version of the model, so the claim is not only about idealized simulation.","feed_headline":"A single Slater determinant suffices for IQPE on the Hubbard hexagon","feed_subtitle":"Noiseless IQPE matches exact ground-state energies across fillings and interaction strengths up to 6.","key_machinery":"The load-bearing object is the pair formed by the initial state and the evolution operator: a single-Slater-determinant initial state prepared through planar rotations, and the IQPE unitary $U=e^{-iHt}$ approximated by a first-order Trotter-Suzuki product with $N_{\\rm trot}=15$ steps, using a Jordan-Wigner mapping that arranges spin-up and spin-down qubits in separate blocks. The key simplification is Eq. (10), which rewrites the periodic boundary hopping term $(a^\\dagger_N a_1 + \\text{h.c.})$ as a local Pauli exchange multiplied by the global phase $(-1)^{N_f-1}$, where $N_f$ is the fermion number in the spin sector; this removes the multi-qubit Jordan-Wigner string from the deepest part of the circuit. IQPE then reads out the phase $\\phi_0$ bit by bit with a single ancilla, and the ground-state energy follows from $E_0=-2\\pi\\phi_0/t$.","core_discovery":"The central claim is that a single Slater determinant carries enough overlap with the true Hubbard ground state that IQPE projects onto it with high probability, so no variational ansatz or state-specific preparation is required for small Hubbard systems. In noiseless simulation with $m=5$ phase bits and $N_{\\rm trot}=15$ Trotter steps, the IQPE energy estimates sit on top of exact diagonalization for every occupation number and interaction strength tested. The mechanism that makes the circuit practical is an identity: for a one-dimensional Hubbard chain with periodic boundary conditions in a fixed particle-number sector, the non-local Jordan-Wigner string attached to the boundary hopping term reduces to the global phase factor $(-1)^{N_f-1}$, removing a long sequence of CNOT gates. The noise analysis then shows that two-qubit depolarizing errors and thermal relaxation corrupt the phase readout most, while readout errors are comparatively harmless, and that halving the hardware error rates moves the energy estimates substantially closer to exact results.","pith_inferences":["The overlap mechanism suggests a quantitative test the paper leaves implicit: the squared overlap between the single Slater determinant and the exact ground state should be computed for every $(N_{\\rm occ}, U_0)$ point, and IQPE's success probability per bit string should track that overlap; if it does not, the apparent convergence may be an artifact of the phase resolution.","If the single-Slater-determinant sufficiency persists, it likely degrades with system size or stronger coupling, so a natural extension is to benchmark IQPE on larger hexagonal flakes or values of $U_0$ beyond 6 to find where the overlap assumption breaks.","The JW-string simplification should transfer to other periodic fermionic chains in fixed-number sectors, so the circuit-depth savings may apply to models such as small Hubbard ladders or spinful chains with periodic boundaries.","Because the noise study shows interaction-dominated circuits are more resilient, a practical error-mitigation strategy would concentrate mitigation effort on the hopping blocks of the Trotter circuit rather than distributing it uniformly."],"forward_implications":["IQPE with a fixed non-variational initial state can serve as a benchmark for ground-state energies of small Hubbard clusters, complementing variational approaches.","The Jordan-Wigner string collapse cuts circuit depth for any one-dimensional fermionic chain with periodic boundary conditions, which directly reduces the two-qubit gate count that the noise study identifies as the main error source.","Noise-aware resource estimates for near-term simulation should budget primarily for two-qubit gate fidelity and coherence time, not for readout correction.","Hardware with roughly half the current two-qubit error rate should bring IQPE energy estimates close to exact values in the weakly interacting regime, according to the paper's halved-noise simulations.","Adiabatic evolution supplies accurate charge and spin densities and correlations, except at fillings $N_{\\rm occ}=4$ and $8$ where the minimum gap vanishes."],"supporting_citations":[{"why":"It supplies the Slater-determinant preparation algorithm used to build the initial states.","marker":"[18]"},{"why":"It sets the QPE-based framework for simulating correlated fermion models, including the Hubbard model, that IQPE builds on.","marker":"[16]"},{"why":"It introduces the single-ancilla iterative phase estimation algorithm that the IQPE implementation follows.","marker":"[39]"},{"why":"It provides the Trotter-Suzuki product formula used to approximate the time-evolution unitary.","marker":"[55]"},{"why":"It establishes the Jordan-Wigner transformation that maps fermionic operators to qubit operators, the basis for Eq. (10).","marker":"[56]"},{"why":"It provides the exact diagonalization results used as benchmarks for the noiseless simulations.","marker":"[57]"},{"why":"It supplies the adiabatic state preparation schedule used in computing densities and correlations.","marker":"[36]"}],"fun_headline_variants":["One Slater determinant seeds IQPE for hexagon Hubbard","Noiseless IQPE nails Hubbard hexagon ground states","Hubbard hexagon: single determinant beats variational fuss","IQPE on graphene hexagon: clean simulations, noisy reality"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the single Slater determinant retains enough overlap with the true ground state at every filling and interaction strength tested; the paper never reports those overlaps and infers sufficiency only from the final energy agreement.","fun_headline_variants_meta":{"raw":{"variants":["One Slater determinant seeds IQPE for hexagon Hubbard","Noiseless IQPE nails Hubbard hexagon ground states","Hubbard hexagon: single determinant beats variational fuss","IQPE on graphene hexagon: clean simulations, noisy reality"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000199,"raw_usage":{"total_tokens":1415,"prompt_tokens":1033,"completion_tokens":382,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":649,"completion_tokens_details":{"reasoning_tokens":314}},"tokens_in":649,"tokens_out":382,"duration_ms":5143,"temperature":1.0,"reasoning_tokens":314,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T10:27:45.901313+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the exact squared overlap between the prepare-and-evolve initial Slater determinant and the exact Hubbard ground state for a point inside the claimed range, say $N_{\\rm occ}=6$, $U_0=6$; if that overlap is essentially zero while noiseless IQPE still returns the exact ground-state energy with high probability on every run, the paper's stated mechanism for convergence would be falsified for that point.","supporting_citations":[{"cited_title":"Somma, G","cited_arxiv_id":null,"evidence_quote":"It supplies the Slater-determinant preparation algorithm used to build the initial states."},{"cited_title":"Schollwöck, The density-matrix renormalization group Rev","cited_arxiv_id":null,"evidence_quote":"It sets the QPE-based framework for simulating correlated fermion models, including the Hubbard model, that IQPE builds on."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It provides the Trotter-Suzuki product formula used to approximate the time-evolution unitary."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It provides the exact diagonalization results used as benchmarks for the noiseless simulations."}],"review_version":1}