{"id":"d6e0ac25-3a43-4a7c-9021-fc82601aaca2","arxiv_id":"1908.00530","paper_version":2,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Self-consistently recalculated occupation numbers in a ladder approximation lower the predicted critical polarization of a polarized Fermi gas at unitarity from 0.834 to 0.543, with the Luttinger-consistent value 0.709.","lead":"A calculation for a gas of atoms with unequal numbers of two spin states now lets the pairing interaction depend on the atoms' own correlated motion. It predicts that pairing disappears at a lower spin imbalance than the standard ladder approximation says, bringing theory closer to quantum Monte Carlo results.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The agreement of r-RPA(∞) with QMC is largely an artifact of Luttinger-theorem violation: the advertised Pc=0.543 corresponds to a real polarization closer to the corrected P_L^c=0.709, so the numerical benchmark comparison is not yet meaningfully established.","rationale":"I read the paper in good faith: it explicitly discloses the Luttinger-theorem violation and provides P_L^c as a diagnostic, so the flaw is not hidden. However, the most load-bearing issue for the central claim is not the second-order/Thouless-criterion caveat, which the paper carefully frames as a word of caution, but rather that the advertised critical polarization in r-RPA(∞) is not the polarization of the system whose occupation numbers are actually computed. The kF-based polarization and the density-based polarization diverge precisely in the regime where the central comparison is made, and the paper never gives the actual density polarization at the unitarity critical point. The P_L^c=0.709 line is a useful consistency correction, but it is not the actual polarization either. The proposed check would settle whether the 0.543 versus 0.562 agreement is physical or an artifact of non-conservation. The qualitative conclusion that r-RPA lowers Pc relative to RPA survives this concern because r-RPA(1st), which satisfies the Luttinger theorem, also gives Pc below RPA, and the paper is otherwise transparent about its pathologies. Therefore the reader's ACCEPT verdict is not overturned, but the strongest quantitative aspect of the claim should be re-expressed after the check is performed.","tokens_in":14112,"tokens_out":12039,"duration_ms":118306,"concrete_test":"At unitarity, run the converged r-RPA(∞) iteration at the nominal critical point k↓F/k↑F = [(1−0.543)/(1+0.543)]^(1/3), compute the actual densities ρσ=∫d³k/(2π)³ nσ(k) from the final occupation numbers, and report P_actual=(ρ↑−ρ↓)/(ρ↑+ρ↓). If P_actual differs from 0.543 by more than about 0.01, re-express the Pc claims in Fig. 12 using P_actual (or, failing that, P_L^c) and re-evaluate the comparison with the ASLDA/QMC value 0.562.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central numerical claim is that r-RPA(∞) gives Pc=0.543 at unitarity, a significant improvement over RPA's 0.834 and close to the ASLDA/QMC value 0.562. But r-RPA(∞) violates the Luttinger theorem, and Table I already shows violations up to 2–3% at moderate coupling, with the text stating that the violation is much worse near the unitary critical point. The quantity plotted as Pc throughout the paper is computed from the input kF values via Eq. (1), not from the actual densities ρσ=∫d³k/(2π)³ nσ(k) of the correlated system. Therefore the ordinate of Fig. 12 is not the physical polarization in the r-RPA(∞) calculation, and the comparison with experiments and ASLDA, which measure density imbalance, is not apples-to-apples. The paper's own Luttinger-consistent estimate, P_L^c=0.709, shows the size of the problem: of the total reduction 0.834−0.543=0.291, only 0.834−0.709=0.125 is attributed to self-consistent occupations once the Luttinger artifact is removed, i.e. roughly 57% of the advertised reduction is a number-conservation artifact at face value. This does not invalidate the qualitative conclusion, since r-RPA(1st) also lowers Pc while satisfying the Luttinger theorem, but it means the headline benchmark agreement rests on an unphysical contribution whose magnitude at Pc is never quantified.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper applies the renormalized particle-particle RPA (r-RPA) to a zero-temperature, spin-imbalanced Fermi gas in the normal phase, on the BCS side of the BCS-BEC crossover and around unitarity. The in-medium T matrix is built from ladder diagrams in which the step-function occupations of standard pp-RPA are replaced by correlated occupation numbers determined self-consistently. Two variants are used: occupations from the truncated Dyson equation [r-RPA(1st)], which satisfy the Luttinger theorem exactly but develop negative quasiparticle weights at strong coupling, and occupations from the resummed Dyson equation [r-RPA(∞)], which remain physical but violate the Luttinger theorem. The authors compute the correlated occupation numbers, the contact, and the critical polarization Pc via the Thouless criterion with the FFLO pairing momentum (Eqs. (22)-(23)). The central finding is that self-consistency lowers Pc relative to standard RPA: at unitarity, Pc = 0.834 (RPA), Pc = 0.543 (r-RPA(∞)), with a Luttinger-consistent estimate P_L^c = 0.709, compared with the ASLDA/QMC benchmark 0.562. The paper also discusses pathologies of the approach in detail: the overlapping two-particle/two-hole continua in the self-consistent J (Fig. 4), the negative-step problem in r-RPA(1st), the Luttinger violation in r-RPA(∞), and the caveat that the Thouless criterion cannot detect a first-order transition with a coexistence region.","tokens_in":14403,"tokens_out":29515,"duration_ms":267727,"significance":"If the central result holds, the r-RPA provides a cheap, physically motivated improvement over bare pp-RPA for imbalanced gases: self-consistently renormalized occupations soften the logarithmic singularity of Re J and thereby push the critical polarization in the right direction relative to independent ASLDA/QMC and experimental estimates. The paper's strengths are its transparency and the concreteness of the formalism: explicit integral expressions for the occupations (Eqs. (9)-(10), (15)-(21)), an unambiguous pole-search procedure, a falsifiable prediction for Pc(1/(a k↑_F)), and unusually direct acknowledgments of the method's failures, including the statement that the whole reduction of Pc in RPA(∞) is a Luttinger artifact, the polaron-to-molecule discrepancy (0.24 vs 0.9), and the warning about first-order transitions. The near-invariance of the contact under self-consistency is also informative.","major_comments":[{"comment":"The paper's headline quantitative claim is that r-RPA(∞) yields Pc = 0.543 at unitarity, “a significant improvement” over RPA (0.834) and close to the ASLDA/QMC value 0.562. But the ordinate of Fig. 12 is the nominal polarization P = (ρ↑ − ρ↓)/(ρ↑ + ρ↓) with ρσ = kσ^3_F/(6π²), i.e., it is evaluated from the input Fermi momenta, not from the actual densities ρσ = ∫ d³k/(2π)³ nσ(k) of the correlated normal state. Table I shows that r-RPA(∞) already violates the Luttinger theorem by 2.1% for the minority spin at the moderate coupling 1/(a k↑_F) = −2.5, and the text states that the violation is “much worse” near the unitary critical point. Hence the physical polarization of the r-RPA(∞) solution at the pole is not 0.543. The paper's own Luttinger-consistent estimate, P_L^c = 0.709, implies that of the total reduction 0.834 − 0.543 = 0.291 only 0.125 survives, i.e., about 57% of the advertised reduction is a number-conservation artifact; the Conclusion's statement that the r-RPA critical polarization is “close to the one of Ref. [38]” holds only for the uncorrected value. I request that the authors either (i) quote Pc computed from the correlated densities at the pole, or (ii) present P_L^c = 0.709 as the meaningful r-RPA(∞) prediction and revise the abstract and Conclusions accordingly.","section":"Sec. III C; Fig. 12; Conclusions"},{"comment":"At unitarity neither r-RPA variant provides a fully physical normal state at the critical point. r-RPA(1st) is in the regime where the minority quasiparticle step Z↓ is negative (dotted red curves in Fig. 12; negative values of n↓ already appear at moderate coupling, cf. Fig. 3(b)), while r-RPA(∞) has a Luttinger violation that the authors say is “much worse” near the unitary critical point than the values in Table I. The manuscript never quantifies the violation, the density integrals, or Z at the critical point, even though all of these are computed in the course of the iterations. Since the unitarity value of Pc is the paper's central quantitative result, I ask for a table or explicit statement reporting Δσ_rel, ρσ = ∫ nσ(k) d³k/(2π)³, and Zσ at the unitary critical point for r-RPA(∞), together with a clear statement that the r-RPA(1st) unitarity value is only obtained in the negative-Z regime.","section":"Sec. III C; Table I; Fig. 12"},{"comment":"The authors show that in the self-consistent calculation the two-particle continuum of Im Jpp leaks below Ω_F and the two-hole continuum of Im Jhh leaks above it, and they call this “a general problem of the r-RPA approach.” Since Re J is obtained from Im J by dispersion, the quantity Re J(kFFLO, Ω_F) entering the Thouless criterion (Eq. (23)) is polluted by these leaked contributions; the pole search in r-RPA therefore does not have the same spectral meaning as in standard RPA. The impact of this pollution on the quoted values of Pc is not assessed. A concrete test would be to recompute Pc with the leaked parts of Im J truncated at Ω_F and compare with the full result; if the difference is small, that should be stated explicitly.","section":"Sec. II B; Fig. 4; Sec. III A"}],"minor_comments":[{"comment":"“writted” should be “written” in “the correlated occupation numbers writted in Eq. (11).”","section":"Sec. II B"},{"comment":"The definition “P_L^c = (k↑^3_F − k↓^3_F)/(k↑^3_F + k↓^3_F)” is formally identical to P of Eq. (1) evaluated at the same Fermi momenta; the authors should clarify that P_L^c is meant to be evaluated with effective Fermi momenta obtained from the correlated densities ρσ = ∫ nσ(k) d³k/(2π)³, since as written the two quantities are indistinguishable.","section":"Sec. III C"},{"comment":"The numerical implementation (momentum grids, cutoffs, convergence criterion for the self-consistent iteration, and the tolerance used in the zero search of Eq. (23)) is not specified; these details are needed for reproducibility.","section":"Secs. II B and III B"},{"comment":"The statement that the converged r-RPA result is independent of the initial occupation numbers is asserted but not demonstrated; a short convergence study would support this, particularly because a special initialization (Z↓ = 0) is needed near Pc.","section":"Sec. III B"},{"comment":"The relations Z1st ≃ 1 + dΣ/dω and Z∞ ≃ 1/(1 − dΣ/dω) are stated without derivation; a two-line derivation would clarify the relation between the two schemes.","section":"Sec. II D"},{"comment":"Fig. 12 mixes theoretical curves for the second-order (Thouless) instability with experimental points for the first-order phase-separation boundary; the caption should state this distinction more prominently, as the body of Sec. III C does.","section":"Fig. 12 caption"}],"recommendation":"major_revision","confidential_remarks":"To the editor: the formal derivation and numerics are sound and the paper is commendably transparent about its limitations; my reservations concern the framing of the central quantitative claim, and they are fully resolvable. The skeptical reading of the headline comparison lands on inspection: the paper's own P_L^c = 0.709 implies that about 57% of the reduction from 0.834 to 0.543 is a Luttinger-theorem artifact, yet the Conclusions state that the r-RPA critical polarization is close to the ASLDA/QMC value, which is only true for the uncorrected number. Because the fixes (reporting density-based polarizations, quantifying the violation at the critical point, testing the effect of the leaked continua, and reframing the abstract and Conclusions) are all within the scope of a revision, I recommend major revision rather than rejection. The reliance on the authors' earlier work (Ref. [20]) is substantive and appropriate, and the novelty of the present results (self-consistent occupations, polarized-gas critical polarization, contact) is clear."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a solid, honest many-body methods paper, and the central qualitative claim—self-consistent occupation numbers in pp-RPA pull the critical polarization down—survives scrutiny. But the specific number the authors advertise at unitarity, Pc = 0.543 from r-RPA(∞), is not the physical polarization. It is computed from input kF values while the calculation violates number conservation, and the Luttinger-consistent value is 0.709. The stress-test note has this right: of the 0.834 → 0.543 reduction, roughly 57% is a Luttinger artifact. The paper says as much, but it still headlines 0.543 as the improvement and compares it to QMC/experimental values that measure density imbalance. That comparison is not apples-to-apples.\n\nWhat is new: this is the first self-consistent r-RPA calculation (both truncated and full Dyson versions) for the zero-temperature polarized Fermi gas, including correlated occupation numbers, FFLO wavevector, and contact. The r-RPA(1st) version satisfies the Luttinger theorem to numerical precision and also lowers Pc, so the qualitative effect is not just an artifact. The discussion of pathologies—negative Z in RPA(1st), spectral continuum mixing, inability to detect first-order transitions—is unusually candid and mostly correct. The FFLO wavevector shift and the polaron-to-molecule point M are useful auxiliary results.\n\nSoft spots, in proportion: the biggest is the reported Pc comparison, as above. Figure 12 labels the ordinate Pc from Eq. (1), not the actual density polarization of the r-RPA(∞) fluid; near the unitary critical point the violation is acknowledged to be “much worse” than Table I’s 2–3%, but never quantified. The claim that convergence is independent of initial conditions is asserted, not demonstrated. Numerical error is not quantified anywhere, and the contact comparison between RPA and r-RPA is consistent only within unknown precision. These are minor except the first.\n\nWho is this for? Cold-atom and nuclear-matter theorists who use ladder approximations. It deserves a serious referee: the method is clearly posed, the limitations are disclosed, and the qualitative conclusion is useful. I would ask the referee to insist on an honest axis in Fig. 12—either use actual densities or restrict claims to the Luttinger-consistent PLc—before publication. That is a major revision concern, but the paper is a genuine contribution.","headline":"A careful, self-disclosed r-RPA implementation that qualitatively lowers the critical polarization of a polarized Fermi gas, but the headline benchmark agreement with QMC rests partly on a Luttinger-theorem violation and should not be taken at face value.","tokens_in":14953,"tokens_out":2805,"would_cite":true,"duration_ms":29630,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper claims that renormalizing particle-particle RPA with self-consistently determined occupation numbers lowers the critical pairing polarization of a polarized Fermi gas, bringing the unitary-limit prediction from 0.834 to 0.543.","keywords":["renormalized RPA","polarized Fermi gas","critical polarization","FFLO phase","BCS-BEC crossover","occupation numbers","Tan contact","Luttinger theorem"],"falsifier":"Measure the onset of pairing in a uniform, flat-trapped imbalanced Fermi gas at unitarity as the polarization is lowered; if superfluidity (or phase separation) appears at a polarization above $P_c=0.543$, the self-consistent softening of the $T$-matrix singularity is overestimated, while a first-order transition found below the predicted pole would show the Thouless criterion misidentifies the boundary.","tokens_in":13843,"feed_emoji":"⚛️","tokens_out":10020,"duration_ms":88296,"temperature":0.7,"pith_summary":"At zero temperature, a spin-imbalanced Fermi gas with more up than down atoms stays superfluid only up to a critical polarization $P_c$. The paper claims that the standard particle-particle RPA (ladder) approach strongly overestimates this $P_c$, and that renormalizing the RPA by feeding self-consistently computed occupation numbers back into the $T$-matrix brings the prediction close to benchmark calculations: at unitarity, $P_c$ drops from 0.834 (RPA) to 0.543 (r-RPA with the full Dyson equation), against a benchmark value of 0.562. The same self-consistent treatment yields correlated occupation numbers and Tan's contact, which is almost unchanged. The paper also quantifies two pathologies: the truncated Dyson equation can produce negative occupation steps in strong coupling, while the full Dyson equation violates the Luttinger theorem, and part of the $P_c$ reduction is attributable to that violation.","feed_headline":"Renormalized RPA drops Fermi-gas critical polarization to 0.543","feed_subtitle":"Standard ladder theory overestimates pairing onset at unitarity; self-consistent occupations line up with QMC benchmarks.","key_machinery":"The central object is the renormalized particle-particle RPA (r-RPA): the in-medium $T$-matrix $\\Gamma(k,\\omega)=1/(1/g-J(k,\\omega))$ is built with occupation numbers $n_\\sigma(k)$ that are themselves computed from the dressed Green's function, either keeping the self-energy to first order (RPA(1st)) or resumming the Dyson equation (RPA($\\infty$)), and iterated until convergence. The two-particle propagator $J(k,\\omega)$ contains particle-particle and hole-hole terms (Eqs. 4 and 5); replacing the Heaviside Fermi steps by correlated occupation numbers reduces the peak of $\\mathrm{Re}\\, J$, which is what delays the appearance of the $T$-matrix pole. The onset of pairing is found by the Thouless criterion, $1/g-\\mathrm{Re}\\, J(k_{\\mathrm{FFLO}},\\Omega_F)=0$, with $k_{\\mathrm{FFLO}}$ the wave vector maximizing this quantity (Eqs. 22 and 23); the quasi-particle step heights $Z_\\sigma$ and the Tan contact $C=\\lim_{k\\to\\infty} k^4 n_\\sigma(k)$ are the diagnostics used to test the scheme.","core_discovery":"The central discovery is that the instability of the normal phase, located by the first pole of the in-medium $T$-matrix (Thouless criterion), depends sensitively on the occupation numbers used in the two-particle propagator. When the occupation numbers are no longer the bare Fermi steps but the correlated ones obtained from the Dyson equation and iterated to self-consistency, the logarithmic singularity in $\\mathrm{Re}\\, J(k,\\omega)$ is softened, so the normal phase remains stable down to lower polarization. On the BCS side and around unitarity this lowers $P_c$ systematically; at unitarity it goes from $P_c=0.834$ in standard RPA to $P_c=0.543$ in the self-consistent version using the full Dyson equation, close to the 0.562 obtained from an energy-density functional fitted to quantum Monte Carlo data. The paper is careful to note that the low value is partly a consequence of the violation of the Luttinger theorem: imposing particle-number conservation would give $P_c=0.709$. The self-consistency also moves the Lifshitz point where the FFLO instability gives way to a $k=0$ pairing to higher polarization asymmetry, and it leaves the contact nearly unchanged.","pith_inferences":["A variant that enforces the Luttinger theorem while using the full Dyson equation would be worth testing; the paper's bracketing values suggest the true improvement over RPA lies between $P_c=0.709$ and $P_c=0.543$ at unitarity.","The same self-consistency loop could be transported to finite temperature, where the standard ladder approach is known to fail for polarized gases; it might reduce the overestimated critical temperature or polarization without the cost of fully self-consistent Green's-function methods.","A flat-trap experiment mapping the onset of pairing versus polarization could test the sharpest signature here: the r-RPA($\\infty$) prediction that the FFLO-type instability only exists for $k_\\downarrow/k_\\uparrow\\gtrsim 0.435$."],"forward_implications":["At unitarity, the r-RPA($\\infty$) critical polarization $P_c=0.543$ is close to the quantum-Monte-Carlo-based value 0.562, a clear improvement over the RPA value 0.834.","Under self-consistency the FFLO-type instability region shrinks: the Lifshitz point moves from $k_\\downarrow/k_\\uparrow=0.224$ (RPA) to 0.435 (r-RPA($\\infty$)), so for stronger pairing the predicted instability switches to a $k=0$ superfluid instead of an oscillating order parameter.","The self-consistent treatment does not cure the negative-step pathology of the truncated Dyson equation; only the full Dyson equation keeps the occupation numbers physical, at the cost of violating the Luttinger theorem.","Tan's contact is almost insensitive to the self-consistent treatment, indicating that the $P_c$ reduction is driven by the softened singularity of $J$, not by a change in the high-momentum tails."],"supporting_citations":[{"why":"Provides the zero-temperature pp-RPA formalism and the RPA occupation-number formulas (Eqs. 9 and 10) that the self-consistency loop recycles.","marker":"[20]"},{"why":"The standard finite-temperature ladder approach whose known failure in polarized gases motivates the self-consistent extension developed here.","marker":"[13]"},{"why":"Introduced renormalized RPA using correlated occupation numbers in many-body theory; this is the basis of the present scheme.","marker":"[25]"},{"why":"Another source of the renormalized-RPA prescription, here extended to polarized Fermi gases.","marker":"[26]"},{"why":"Thouless criterion: the pole of the $T$-matrix at $\\omega=\\Omega_F$ locates the pairing instability used to define $P_c$.","marker":"[36]"},{"why":"Energy-density functional fitted to quantum Monte Carlo results; supplies the benchmark unitary-limit critical polarization 0.562.","marker":"[38]"},{"why":"Review with the mean-field phase diagram and FFLO Lifshitz-point reference used for comparison with the r-RPA results.","marker":"[34]"},{"why":"Luttinger theorem; defines the exact particle-number relation whose violation in RPA($\\infty$) is quantified in Table I.","marker":"[35]"}],"fun_headline_variants":[],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The critical polarization is located by the first pole of the $T$-matrix at the FFLO wave vector (Thouless criterion), which assumes the normal-to-superfluid transition is a second-order instability; if the true transition is first order with a coexistence region extending to higher polarization, the calculation cannot detect it and the quoted $P_c$ is not necessarily the physical phase boundary.","fun_headline_variants_meta":{"error":"Client error '402 Payment Required' for url 'https://api.deepseek.com/chat/completions'\nFor more information check: https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/402"},"cache_creation_input_tokens":0},"created_at":"2026-08-14T15:49:37.239806+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the onset of pairing in a uniform, flat-trapped imbalanced Fermi gas at unitarity as the polarization is lowered; if superfluidity (or phase separation) appears at a polarization above $P_c=0.543$, the self-consistent softening of the $T$-matrix singularity is overestimated, while a first-order transition found below the predicted pole would show the Thouless criterion misidentifies the boundary.","supporting_citations":[{"cited_title":"Urban and P","cited_arxiv_id":null,"evidence_quote":"Provides the zero-temperature pp-RPA formalism and the RPA occupation-number formulas (Eqs. 9 and 10) that the self-consistency loop recycles."},{"cited_title":"Nozi` eres and S","cited_arxiv_id":null,"evidence_quote":"The standard finite-temperature ladder approach whose known failure in polarized gases motivates the self-consistent extension developed here."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Another source of the renormalized-RPA prescription, here extended to polarized Fermi gases."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Thouless criterion: the pole of the $T$-matrix at $\\omega=\\Omega_F$ locates the pairing instability used to define $P_c$."},{"cited_title":"Bulgac and M","cited_arxiv_id":null,"evidence_quote":"Energy-density functional fitted to quantum Monte Carlo results; supplies the benchmark unitary-limit critical polarization 0.562."},{"cited_title":"Calvanese Strinati, P","cited_arxiv_id":null,"evidence_quote":"Review with the mean-field phase diagram and FFLO Lifshitz-point reference used for comparison with the r-RPA results."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Luttinger theorem; defines the exact particle-number relation whose violation in RPA($\\infty$) is quantified in Table I."}],"review_version":1}