{"id":"66b2522f-256f-4d6d-ae3d-0cdd167e08a6","arxiv_id":"1908.05606","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A new 'even-odd' self-consistent RPA scheme derives single-particle occupancies from the odd-particle channel and reproduces exact solutions of the Lipkin and Hubbard models.","lead":"This paper couples the equations for even and odd particle numbers in self-consistent RPA theory, using the odd channel to determine single-particle occupation numbers. It tests the method on the Lipkin and one-dimensional Hubbard models and finds close agreement with exact solutions.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The admitted failure of Q|Z>=0 (Eq. 8) is the load-bearing uncontrolled approximation; a quantitative check of the residual is needed before the eo-SCRPA scheme can be considered predictive.","rationale":"The reader's weakest-assumption analysis identifies Eq. (8), the assertion that the even-RPA destruction operator Q exactly annihilates the correlated ground state, as the load-bearing approximation. This is precisely the point on which the paper is least secure: the text itself concedes that Q does not kill the ground state, and all subsequent evaluations of correlation functions rely on commuting Q to the vacuum. No error bound or quantitative estimate of the residual Q|Z> is provided. The numerical agreement with exact results in the Lipkin and Hubbard models is genuine evidence that the approximation is benign in those two small test cases, but it does not by itself establish the absence of a hidden inconsistency, because the self-consistent equations could in principle converge to a solution while the underlying killing condition is violated. The proposed test—computing the normalized norm of Q|Z> and comparing exact versus commuted correlation functions—directly probes the internal consistency of the scheme. If the residual is small and the correlation functions agree, the concern is resolved for those models; if not, the central claim loses support. Since the reader already reached a CONDITIONAL verdict and my stress test identifies the same concern without finding a new fatal issue, the appropriate outcome is to leave the verdict unchanged.","tokens_in":14481,"tokens_out":8312,"duration_ms":84937,"concrete_test":"For the Lipkin model with N=10, take the self-consistent eo-SCRPA solution at several couplings (e.g., chi = 0.5, 1.0, 1.5). Construct the CCD-type state |Z> using the Z amplitudes consistent with Eq. (4), and the RPA operator Q from the converged X,Y amplitudes. Compute the normalized residual R = <Z|Q^dagger Q|Z>/<Z|Z>. Then compare the exact value of a two-body correlator, such as <Z|J+J-|Z>, with the value obtained by commuting Q to the vacuum using Eq. (8), as done in the paper. If R is not small (say, >0.01) or if the two evaluations of <Z|J+J-|Z> differ by more than 10%, the central approximation is quantitatively uncontrolled and the good agreement in the tested models is not a reliable indicator of predictive power.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central new ingredient of the paper is the closed set of coupled equations (9) and (14), with all correlation functions evaluated by commuting the even-RPA destruction operator Q to the right until it acts on the correlated vacuum |Z>. The paper explicitly states in Section II, after Eq. (7), that 'the even particle number equation (SCRPA) is the only point where the approach is not entirely consistent' and that Q does not exactly kill the CCD ground state, yet it then imposes Eq. (8), Q|Z>=0, and uses this to evaluate every correlator entering A, B, C, and D. No estimate is given of the magnitude of Q|Z> or of the error this induces in the two- and three-body correlation functions. Because the occupancies from Eqs. (18)-(19) and the X,Y amplitudes from Eq. (9) feed back into each other, a small violation of Eq. (8) could be amplified by the self-consistency loop. The agreement with exact results in the Lipkin and Hubbard models is encouraging, but it does not establish the regime of validity of the method: the two test models are small, and the coupled equations contain additional uncontrolled truncations (e.g., the three-body terms in D are approximated by keeping only diagonal density insertions). The load-bearing assumption is therefore Eq. (8), whose failure is admitted but never quantified.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a coupled self-consistent RPA scheme, called eo-SCRPA, in which the even-particle SCRPA equations for the correlated ground state and an odd-particle RPA equation for single-particle motion are solved self-consistently from the same correlated vacuum. The new ingredient is the calculation of single-particle occupation numbers from the odd-RPA amplitudes, Eqs. (18)--(19). The authors derive the general coupled equations (9) and (14), evaluate all correlation functions by commuting the even-RPA destruction operator Q to the vacuum using the assumption Q|Z>=0, and then test the scheme on the Lipkin model and a six-site Hubbard ring at half filling. The reported occupancies, excitation energies, and ground-state energies agree well with exact diagonalization across a range of couplings, with the Lipkin N=4 case reproduced exactly.","tokens_in":14817,"tokens_out":6430,"duration_ms":65466,"significance":"If the scheme is valid, it provides a parameter-free method to obtain correlated occupation numbers within the SCRPA framework, addressing a long-standing closure problem. The paper contains detailed derivations and benchmarks against two exact solvable models, with machine-checkable equations and no fitted parameters. The numerical agreement is encouraging and goes beyond earlier Catara-method results. However, the theoretical foundation rests on the uncontrolled assumption Q|Z>=0, which the authors themselves state is not exactly fulfilled, and on additional truncations in the three-body matrix D. A quantitative estimate of these errors is needed before the method can be regarded as a reliable predictive tool.","major_comments":[{"comment":"The assumption Q|Z>=0 is load-bearing: all correlation functions entering A, B, C, and D are evaluated by commuting Q to the right and dropping Q|Z>, yet the text explicitly states that Q does not exactly kill the CCD ground state. No estimate of ||Q|Z>|| or of the induced error in the two- and three-body correlations is given. Because the self-consistency loop couples occupancies (18)--(19) with the amplitudes of (9) and (14), a small residual could be amplified. I request a quantitative check in the solvable models, for example by computing <Z|Q†Q|Z>/<Z|Z> from the converged SCRPA solution, or by comparing the occupancies obtained with and without the killing condition.","section":"Sec. II, Eq. (8)"},{"comment":"The three-body matrix D is evaluated by retaining only those terms where a particle state of the triplet operator connects to the interaction, with the remaining density operator taken diagonal. The authors justify the discarded terms as 'supposedly less important' and 'probably small,' but no measure of their magnitude is provided. This truncation is part of the equations solved in the applications, so its accuracy is load-bearing for the numerical agreement. I ask for a comparison of the full D, or at least an estimate of the neglected terms, in the small test systems.","section":"Sec. II, Eq. (14) and Fig. 1"},{"comment":"The main text states that <J0J0> is factorized as <J0>^2, while Appendix A also offers the Casimir relation (A7) for the same quantity. It is not clear which version was used to produce Figs. 2--7. If the factorization is used, the results depend on an additional uncontrolled approximation; if the Casimir relation is used, the presentation should say so. Please clarify and, if feasible, show the sensitivity of the results to this choice.","section":"Sec. III.A, after Eq. (A6)"},{"comment":"The derivation of the single-particle occupation numbers from the odd-RPA amplitudes assumes that {a†_h, q†_{h,\\mu}} = x^\\mu_h with all U terms vanishing, and that the odd operators form a complete set for one-hole excitations. Neither a completeness relation analogous to Eq. (17) nor a derivation of the anticommutator simplification is provided. Without this, the central new expression for n_h and n_p is an assumption. I request either a derivation or a reference establishing the completeness of the odd-RPA operators.","section":"Sec. II, Eqs. (18)--(19)"}],"minor_comments":[{"comment":"There are several typos, including 'Hatree Fock' in the Fig. 8 caption, 'perfoment' in Sec. II, 'full-fill' in Eq. (4), and 'the the' in Sec. II ('We should also say the the formal expressions').","section":"Throughout"},{"comment":"The two commutator equations for [Q_\\nu, \\hat n_k] are written with identical expressions except for a sign; one of them should presumably be for the Hermitian conjugate or for a different index. Please correct.","section":"Eq. (B9)"},{"comment":"The first term on the right-hand side appears to be missing the \\nu' dependence: it should be X_i^\\nu X_i^{\\nu'} - Y_i^\\nu Y_i^{\\nu'}, not X_i X_i - Y_i Y_i, unless a delta function is intended.","section":"Eq. (B10)"},{"comment":"The inversion should specify that the sum over \\nu runs over all RPA modes and that the completeness of the X,Y amplitudes is required for the relation to hold.","section":"Eq. (17)"},{"comment":"The conclusion states that the equations 'should be solvable with modern computers for realistic problems' but provides no complexity estimate; a brief statement of the scaling with system size would be useful.","section":"Sec. IV"}],"recommendation":"major_revision","confidential_remarks":"The paper is within scope for a nuclear theory / many-body methods journal. No citation concerns. The main revision items are the missing error estimate for Eq. (8) and the insufficiently justified completeness assumption behind Eqs. (18)--(19); these are central to the method's validity and require additional numerical or analytic work."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This paper does something concrete: it couples the even-particle-number SCRPA equations with an odd-particle-number RPA equation, originally from Tohyama and Schuck, and uses the odd eigenvectors to produce self-consistent single-particle occupancies. That coupling is new, and it is a sensible way to close the SCRPA equations. The two test cases—Lipkin and 6-site Hubbard—are small but exactly solvable, and the agreement with exact diagonalization is genuinely good, especially for the Lipkin model where the method beats the Catara occupancy scheme. The derivation is laid out in enough detail to reimplement, including the diagrams for the three-body terms and the appendix formulas.\n\nThe soft spot is the same one the reader flagged. The whole machinery uses Q|Z>=0 as an exact identity to evaluate every correlation function, while the text admits in Section II that Q does not exactly annihilate the CCD ground state. That is a real inconsistency, and it is load-bearing: occupancies, X and Y amplitudes, and the odd-RPA norms all feed into each other through the self-consistency loop. A quantitative check of the residual, such as an estimate of <Z|Q+Q|Z> or a comparison against a higher-order truncation, would have made the method far more convincing. The paper also keeps only diagonal density insertions in the three-body block D and factorizes <J0^2> in the Lipkin derivation; these are milder, but they add to the uncontrolled side of the scheme. The absence of code or numerical data files means the results are not independently reproducible, though the equations are explicit enough for a determined reader to implement.\n\nNone of this kills the paper for me. It is an honest methods paper: the central approximation is stated, not hidden, and the benchmarks—especially the lambda+ and lambda- curves and the N=4 exact case—show the method deserves attention. What is missing is a regime-of-validity estimate, not a miracle.\n\nI would send this to peer review. A good referee will ask for the residual check and a third application, but the idea is solid and the presentation is clear.","headline":"A genuinely new coupling scheme for self-consistent occupancies, honestly presented and well benchmarked, though the admitted Q|Z>=0 approximation remains unquantified.","tokens_in":15254,"tokens_out":1645,"would_cite":false,"duration_ms":16505,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["21.60.-n","21.60.Fw","71.10.-w","75.10.Jm"],"model":"deepseek-v4-flash","headline":"This paper claims that single-particle occupation probabilities can be obtained self-consistently from an odd-number RPA equation coupled to the even SCRPA equations, and that the coupled system reproduces exact solutions of the Lipkin…","keywords":["self-consistent RPA","odd-particle-number RPA","single-particle occupation numbers","equation of motion method","Lipkin model","Hubbard model","correlated ground state","Green's function"],"falsifier":"For the Lipkin model at, say, $N=20$ and coupling near $\\chi=1$, take the exact ground state and the $Q$ operator from the converged eo-SCRPA solution and compute $\\langle Z|Q^\\dagger Q|Z\\rangle$; if this norm is not far below one, the commuting-to-vacuum evaluation underlying equations (18)--(19) is invalid and the reported accuracy is accidental. A similar check on the Hubbard ring would compare eo-SCRPA occupancies with the exact reduced density matrices.","tokens_in":14265,"feed_emoji":"⚛️","tokens_out":6002,"duration_ms":57775,"temperature":0.7,"pith_summary":"This paper tries to close the self-consistent RPA scheme by deriving single-particle occupation probabilities from the same equation-of-motion framework that gives the RPA itself, rather than fixing them by a separate ad hoc rule. It sets up two coupled eigenvalue problems: the even-particle-number SCRPA equation for collective excitations and an odd-particle-number RPA equation whose eigenvectors directly yield the occupancies. Because both are built on one correlated ground state, solving them together determines all correlation functions needed in the theory. If the approach is right, strongly correlated systems such as nuclei and lattice electrons can be described without external occupation input, with even and odd excitation spectra obtained from one calculation. Tests on exactly solvable Lipkin and 1D Hubbard models support the scheme.","feed_headline":"Odd-particle RPA sets occupations and matches exact models","feed_subtitle":"Even and odd RPA share one vacuum, so occupancies come from inside the calculation, not from an ad hoc rule.","key_machinery":"The central object is the coupled 'eo-SCRPA' system: the even SCRPA eigenvalue problem (9), whose amplitudes $X,Y$ describe collective particle-hole excitations, and the odd RPA problem (14), whose eigenvectors $x^\\mu_h$, $x^\\rho_p$ are the weights of adding or removing one particle. The load-bearing link is the inversion (17), which rewrites the particle-hole operators of the Hamiltonian in terms of the RPA phonons $Q^\\dagger$, and the killing condition (8), $Q|Z\\rangle=0$, which lets every correlation function be evaluated by commuting $Q$ to the vacuum until it annihilates the correlated ground state. Occupation numbers then reduce to the simple sums (18)--(19) of odd-RPA amplitudes, closing the nonlinear system.","core_discovery":"The central claim is that the single-particle occupation numbers that enter self-consistent RPA need not be supplied externally: they emerge from a companion RPA eigenvalue equation built from odd-particle-number excitations, and the two equation sets are coupled because both are evaluated in the same correlated ground state. Concretely, the even SCRPA equation (9) determines the RPA amplitudes $X,Y$, while the odd equation (14) determines amplitudes $x,U$ whose squared components give the hole and particle occupancies through (18)--(19). The paper reports that in the Lipkin model and the six-site half-filled Hubbard ring this coupled scheme reproduces exact occupancies, excitation energies, and ground-state energies very well; in the Lipkin model it is exact for $N=4$ and improves on earlier ways of fixing occupancies.","pith_inferences":["Because the odd equation can be rewritten as a Dyson equation for the one-particle Green's function, a natural extension the paper does not report is momentum-resolved spectral functions for the Hubbard chain; those are already implied by the self-consistent occupancies and excitation poles.","The systematically larger gain in the Lipkin model than in the Hubbard ring suggests that the main error source is the over-completeness or Pauli violation of the collective phonon basis; measuring the norm of $Q|Z\\rangle$ on exact ground states of larger Lipkin samples would predict where the method degrades.","The authors call the exact $N=4$ Lipkin result a lucky accident; if the same exactness appears for other small integrable systems such as pairing models, it may reflect a hidden constraint rather than chance."],"forward_implications":["The same calculation yields even excitation energies, odd addition and removal energies ($\\lambda_\\pm$), occupation numbers, and the ground-state energy through the one-particle Green's function, so a single eo-SCRPA run replaces several separate approximations.","In the Lipkin model eo-SCRPA gives the exact result for $N=4$ and tracks the exact solution up to and slightly beyond the critical coupling, where standard RPA breaks down.","In the six-site half-filled Hubbard ring, occupation numbers, excitation energies, and ground-state energy agree well with exact results for interaction strengths up to about $U/t \\sim 3$.","The coupled equations are nonlinear in occupancies and in the $X,Y$ amplitudes, so the method requires no external occupation input; the only input is the Hamiltonian and the mean-field basis."],"supporting_citations":[{"why":"Supplies the SCRPA equations, the inversion of the RPA operator, and the model studies showing $Q$ approximately kills the correlated vacuum.","marker":"[3]"},{"why":"Introduces the odd-particle-number RPA equation used here as the starting point for self-consistent occupations.","marker":"[9]"},{"why":"Is the standard Catara method for occupancies that the new odd-RPA occupations replace and improve upon.","marker":"[7]"},{"why":"Provides the earlier SCRPA Hubbard-model application and the benchmark formulas this paper extends.","marker":"[8]"},{"why":"Gives the standard RPA formulation and the Lipkin model setup used throughout.","marker":"[10]"},{"why":"Supplies the Green's function formulas used to obtain the ground-state energy from the self-consistent occupancies.","marker":"[12]"}],"fun_headline_variants":["Self-consistent RPA derives occupancies from odd-number equations","Coupled even-odd RPA yields exact occupancies in test models","Odd-particle RPA sets occupancies without external input","Occupation numbers emerge from coupled RPA equations"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the even-RPA destruction operator $Q$ exactly annihilates the correlated ground state, $Q|Z\\rangle=0$, meaning the ground state contains no RPA excitations; the paper states this is only approximately true, yet every correlation function in the coupled equations is evaluated by commuting $Q$ to the vacuum, and if the residual is not small the whole self-consistency loop is uncontrolled.","fun_headline_variants_meta":{"raw":{"variants":["Self-consistent RPA derives occupancies from odd-number equations","Coupled even-odd RPA yields exact occupancies in test models","Odd-particle RPA sets occupancies without external input","Occupation numbers emerge from coupled RPA equations"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00021,"raw_usage":{"total_tokens":1343,"prompt_tokens":809,"completion_tokens":534,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":425,"completion_tokens_details":{"reasoning_tokens":467}},"tokens_in":425,"tokens_out":534,"duration_ms":5687,"temperature":1.0,"reasoning_tokens":467,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:24:22.782241+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"For the Lipkin model at, say, $N=20$ and coupling near $\\chi=1$, take the exact ground state and the $Q$ operator from the converged eo-SCRPA solution and compute $\\langle Z|Q^\\dagger Q|Z\\rangle$; if this norm is not far below one, the commuting-to-vacuum evaluation underlying equations (18)--(19) is invalid and the reported accuracy is accidental. A similar check on the Hubbard ring would compare eo-SCRPA occupancies with the exact reduced density matrices.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the SCRPA equations, the inversion of the RPA operator, and the model studies showing $Q$ approximately kills the correlated vacuum."},{"cited_title":"Jemai, P","cited_arxiv_id":null,"evidence_quote":"Introduces the odd-particle-number RPA equation used here as the starting point for self-consistent occupations."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Is the standard Catara method for occupancies that the new odd-RPA occupations replace and improve upon."},{"cited_title":"Tohyama, P","cited_arxiv_id":null,"evidence_quote":"Gives the standard RPA formulation and the Lipkin model setup used throughout."},{"cited_title":"Jemai and P","cited_arxiv_id":null,"evidence_quote":"Supplies the Green's function formulas used to obtain the ground-state energy from the self-consistent occupancies."}],"review_version":1}