{"id":"3895446f-0155-4037-8934-495937e7eb2c","arxiv_id":"2602.06298","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"TPSC+ calculations show that the doping at which the antinodal spin-density-wave precursor crosses zero energy coincides with the maximum of the compressibility and of the Knight shift in the 2D Hubbard model.","lead":"This paper uses an approximate many-body method (TPSC+) to show that the compressibility maximum of the two-dimensional Hubbard model as a function of doping occurs just as the lower spin-density-wave precursor band crosses zero energy at the antinodal point. It explains recent cold-atom observations and predicts a similar maximum in the Knight shift, offering a second experimental probe of the pseudogap-to-metal crossover.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"TPSC+ underestimates spin correlations by ~30% at U=7, so the claimed δ_max / precursor-crossing coincidence is established only within an approximation whose quantitative error is acknowledged and untested against exact results in that regime.","rationale":"I read the paper as making a specific, falsifiable claim: the doping maximum of the compressibility and the uniform spin susceptibility are controlled by the same single-particle crossover in which the antinodal SDW precursor crosses the Fermi level, and this happens for weak and intermediate U via thermal SDW fluctuations. The paper is internally coherent and provides a plausible mechanism, with quantitative support at U=3.69 against cold-atom data and thermodynamic-limit checks at L≥64. Credit is due for the finite-size analysis and for explicitly comparing ∂n/∂μ with first-level χ_ch(0,0). However, the headline quantitative coincidence at U=7 rests entirely on TPSC+, and the paper contains its own admission that TPSC+ underestimates spin correlations by 27–33% and underestimates pseudogap strength in exactly this regime. Because the proposed mechanism is spin-fluctuation driven, that level of error could shift the precursor crossing and δ_max, or even change their ordering. The internal inconsistency documented in Sec. VI, where the low-temperature compressibility maximum from the second-level self-energy is not reproduced by the first-level charge susceptibility, further weakens the inference: an exact theory must satisfy ∂n/∂μ=χ_ch(0,0), and TPSC+ does not at low T. The paper acknowledges this and argues the second-level result is better, but that argument is not independently validated. For these reasons the central claim should remain conditional until an exact or systematically controlled benchmark at U≈7, T≈0.07 confirms the maximum and its coincidence with the spectral crossing. This does not change the reader's verdict, but it sharpens the condition: the decisive test is an exact finite-temperature calculation, not further TPSC+ self-consistency.","tokens_in":35522,"tokens_out":4736,"duration_ms":51150,"concrete_test":"Perform DQMC (or DiagMC where sign permits) for U=7, T=1/14 on the largest accessible lattice, e.g., L=8 or L=10, compute κ(δ)=∂n/∂μ and χ_sp(0,0)(δ) by numerical differentiation / correlation functions, and locate their maxima. From the same Monte Carlo Green's function, using maximum-entropy or stochastic analytic continuation, locate the antinodal k=(π,0) lower-band spectral-weight crossing of ω=0. If the exact δ_max and the exact crossing agree within ~0.02 and both lie near 0.15–0.16, the concern is resolved; if either quantity lacks a maximum or shifts by more than ~0.05, the central coincidence fails.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that the maximum in κ(δ) at δ_max≈0.15 coincides with the doping δ≈0.16 where the antinodal lower-SDW precursor crosses ω=0 (Sec. V A, Figs. 6–7, U=7, T=0.0714). This is a quantitative statement about the Hubbard model, but the only evidence is TPSC+. The paper itself reports in Sec. III C that at U=7 TPSC+ underestimates equal-time spin correlations by 27–33%, and Sec. II A states that for U≳3.5 TPSC+ underestimates pseudogap effects, matching exact results only at lower T or lower doping. Since the proposed mechanism is SDW fluctuations, a ~30% underestimate of spin correlations should shift the doping/temperature at which the precursor actually crosses zero and could shift δ_max materially. The coincidence between δ_max and the crossing could survive, but only if the error is nearly uniform in doping; that is not demonstrated. Additionally, Sec. VI shows that at T=0.025 the compressibility ∂n/∂μ computed with the second-level self-energy and χ_ch(0,0) computed with constant U_ch disagree qualitatively in the pseudogap region, so the low-temperature maximum in κ(δ) is not corroborated by a conserving charge response. If a similar inconsistency persists at T=0.0714, the maximum—and its coincidence with the spectral crossing—could be an artifact of the second-level self-energy rather than a robust physical feature. The abstract's additional claim of agreement with the cold-atom Knight-shift temperature maximum is also not demonstrated in the body, but the more load-bearing issue is the unvalidated quantitative accuracy of TPSC+ at the parameters of the headline result.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper uses the TPSC+ approach for the nearest-neighbor square-lattice Hubbard model to study the doping-driven crossover from a pseudogapped electronic liquid to a correlated Fermi liquid. It reports that the isothermal compressibility κ(δ) has a maximum that, for U=7 and T=1/14, occurs at δ_max≈0.15, practically coinciding with the doping δ≈0.16 at which the lower antinodal SDW precursor crosses ω=0 (Sec. V A, Figs. 6–7). The same mechanism is claimed to produce a maximum in the uniform spin susceptibility χ_sp(0,0)(δ) at comparable doping, and both maxima are predicted for weak and intermediate U. The paper also discusses finite-size effects, incommensurate spin fluctuations at low T, and a breakdown of the equivalence between ∂n/∂μ and χ_ch(0,0) within TPSC+ (Sec. VI).","tokens_in":35905,"tokens_out":4074,"duration_ms":41504,"significance":"If the central coincidence is robust, the paper provides a concrete microscopic mechanism connecting a thermodynamic anomaly (the compressibility maximum) to the reconstruction of the single-particle spectrum, and it makes a falsifiable prediction for the Knight shift maximum versus doping. The calculations are internally consistent, use sum-rule-fixed vertices rather than fitted parameters, and reproduce the cold-atom compressibility quantitatively at U=3.69. The finite-size analysis and the explicit discussion of the method's limitations are also valuable. However, the quantitative coincidence at U≈7 rests entirely on an approximation that the paper itself reports underestimates spin correlations by 27–33% at that interaction, so the robustness of the central claim is not yet established.","major_comments":[{"comment":"The central claim, δ_max≈0.15 with the antinodal precursor crossing at δ=0.16, is a TPSC+ result at U=7, T=0.0714. At the same U, Sec. III C reports that TPSC+ underestimates equal-time spin correlations by 27–33%. Since the proposed mechanism is SDW fluctuations, a doping-dependent error of this size could shift both δ_max and the crossing doping. The near coincidence could survive if the error is uniform in doping, but that is not demonstrated. A robustness test—for example, a comparison with DiagMC or DQMC near T≈0.07, or a controlled sensitivity study in U_sp/χ_sp—is needed before the coincidence can be stated as more than an internally consistent TPSC+ result.","section":"Sec. V A, Figs. 6–7"},{"comment":"The paper shows that TPSC+ does not satisfy the exact identity κ=∂n/∂μ=χ_ch(0,0). At T=0.025, the first-level χ_ch(0,0)(δ) is monotonic and shows no maximum, while κ(δ) does; at T=0.15 neither shows a well-defined maximum. No comparison is shown at the T=0.0714 used for the central claim. This leaves open the possibility that the maximum in κ(δ) and its coincidence with the spectral crossing are artifacts of the second-level self-energy rather than a robust property. The manuscript should either provide a consistency check at T≈0.0714 or explicitly justify why the second-level result is the reliable one in that regime.","section":"Sec. VI, Fig. 14"},{"comment":"The abstract states that TPSC+ correctly predicts a maximum in the temperature dependence of the Knight shift, χ_sp(0,0)(T), consistent with cold-atom experiments (Ref. [45]). The body of the paper contains no such calculation or quantitative comparison; the only χ_sp results are versus doping. This claim is currently unsupported and should either be backed by a dedicated figure and comparison or removed from the abstract and conclusions.","section":"Abstract"}],"minor_comments":[{"comment":"The text near Fig. 7 says the U=3.69 case has \"a smaller δ_max = 0.15\", but earlier in the same section and in Figs. 4–5 δ_max=0.1 for U=3.69. This appears to be a typo and should be corrected.","section":"Sec. V A"},{"comment":"The caption lists k=(3π/4, π/4) while the main text refers to k=(3π/8, π/8) for the same panel. Please make the notation consistent.","section":"Fig. 9 caption"},{"comment":"There are several typos in the conclusion, e.g., \"χ_sp(,0,0)\" on two occasions. These should be cleaned up.","section":"Sec. IX"},{"comment":"The caption says \"The doping-dependent isothermal compressibility κ(δ) shows a maximum for T=0.1\", but the figure also includes T=0.15, where no maximum is visible. Clarify which curves show maxima.","section":"Fig. 4 caption"},{"comment":"The factor \\tilde g_{fb} is used in Eq. (14) before its values are introduced later in Sec. II B 2. A brief definition at first use would improve readability.","section":"Eq. (14)"}],"recommendation":"major_revision","confidential_remarks":"The paper is honest about its limitations and the central derivation is internally consistent. The main risk is that the quantitative coincidence at U=7 is only tested within one approximate method with a known ~30% spin-correlation error in exactly that regime. The abstract also overreaches relative to the body on the χ_sp(T) claim. I would not reject, but the authors should provide an explicit robustness check or substantially soften the claims."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper does a genuinely useful thing: it takes the cold-atom compressibility maximum and gives it a concrete single-particle mechanism—the doping at which the lower SDW precursor at the antinodal point crosses zero frequency. The TPSC+ fits to the U=3.69 cold-atom data are quantitative and convincing, and the prediction that the Knight shift should show a maximum as a function of doping is a clean, falsifiable statement. The authors are also unusually honest about their method's limits, which makes the paper easy to read critically.\n\nThe main soft spot is the U=7 coincidence. The paper reports δ_max=0.15 and the antinodal precursor crossing at δ=0.16, but at that interaction TPSC+ underestimates spin correlations by 27–33% and underestimates pseudogap effects. That means the coincidence is only established within an approximation whose quantitative error is acknowledged and untested against exact results in that regime. It could survive—if the error is roughly uniform in doping—but that is not demonstrated. I'd like the referees to ask the authors to show how δ_max and the crossing move when the spin vertex is adjusted to match the exact correlation function, even at a few points.\n\nSecond, the abstract claims agreement with the cold-atom Knight-shift temperature maximum, but the body only shows χ_sp(0,0)(δ). I could not find the χ_sp(0,0)(T) comparison. That claim needs to be either shown or removed.\n\nThird, the Sec. VI comparison between ∂n/∂μ and χ_ch(0,0) shows a qualitative inconsistency at T=0.025 in the pseudogap region. The headline coincidence is at T=0.0714, and the paper does not show that the inconsistency is small there. This is not a fatal flaw, but it means the low-temperature compressibility maximum is not corroborated by a conserving charge response.\n\nNone of this kills the paper. The weak-coupling result is grounded in experiment, the mechanism is plausible, and the Knight-shift prediction is worth testing. The paper is well-cited on prior work—they credit CDMFT and iPEPS appropriately, and the self-citations are to their own established TPSC+ method, not a problem.\n\nI'd send this to peer review. A good referee should push on the U=7 coincidence, the missing Knight-shift temperature comparison, and whether the charge-response inconsistency affects the headline temperature. It's a serious paper from serious people, and the community would benefit from it being out there with those caveats made explicit.","headline":"A solid TPSC+ study that ties the cold-atom compressibility maximum to the antinodal SDW-precursor crossing and predicts a Knight-shift maximum, but the headline coincidence at U=7 rests on an approximation that underestimates spin correlations by ~30%.","tokens_in":36432,"tokens_out":2350,"would_cite":false,"duration_ms":24976,"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":"Maximum of the compressibility in the 2D Hubbard model marks the crossover from a pseudogapped electronic liquid to a correlated Fermi liquid, and practically coincides with the doping where the antinodal spin-density-wave precursor crosses","keywords":["pseudogap","compressibility maximum","Knight shift","spin-density-wave fluctuations","two-dimensional Hubbard model","TPSC+","Van Hove singularity","incommensurate spin fluctuations"],"falsifier":"A determinant or diagrammatic quantum Monte Carlo calculation at U=7 and T=0.0714 in the thermodynamic limit, locating both the maximum of κ(δ) and the doping at which the antinodal SDW precursor crosses ω=0; if these two dopings differ beyond numerical uncertainty, or if χ_sp(0,0)(δ) has no maximum at T=0.1, the central claim is refuted.","tokens_in":1545,"feed_emoji":"🧲","tokens_out":2610,"duration_ms":80642,"temperature":0.7,"pith_summary":"This paper argues that the doping at which the compressibility of the two-dimensional Hubbard model reaches a maximum is the thermodynamic fingerprint of the crossover from a pseudogapped electronic liquid to a correlated Fermi liquid. Using a non-perturbative approximation called TPSC+, it shows that this maximum nearly coincides with the doping at which the precursor of the lower spin-density-wave band at the antinodal point (π,0) crosses zero frequency, shifting the Van Hove singularity in the density of states from occupied to unoccupied energies. The paper also predicts that the uniform spin susceptibility (the Knight shift) has a maximum at the same doping at low temperature, and that both maxima appear for both weak and intermediate interaction strengths, driven by thermally excited critical spin fluctuations with a correlation length of only a few lattice spacings. A sympathetic reader would care because this explains recent cold-atom observations of a compressibility peak and connects a thermodynamic anomaly to a specific single-particle spectral event.","feed_headline":"Compressibility maximum marks the pseudogap-to-Fermi-liquid crossover","feed_subtitle":"A Hubbard-model calculation ties the peak to the spin-density-wave precursor crossing zero, and predicts a Knight-shift maximum.","key_machinery":"The central object is the precursor of the lower (π,π) spin-density-wave band at the antinodal point k=(π,0), a peak in the spectral function below the Fermi level whose position relative to ω=0 changes with doping. The identity that carries the argument is δ_max ≈ δ_{ω=0}: the doping of the compressibility and spin-susceptibility maxima is (almost) the doping at which this precursor crosses zero. The machinery producing the precursor is the TPSC+ self-energy built from static classical spin fluctuations, with two analytical criteria: at regular Fermi-surface points a pseudogap opens when the correlation length ξ exceeds v_F/(πT), while at the antinodal Van Hove point the easier condition ξ>","core_discovery":"On the paper's own terms: for the nearest-neighbor square-lattice Hubbard model, TPSC+ predicts that at low temperature the doping-dependent compressibility κ(δ) has a maximum at δ_max, and this maximum is located at essentially the same doping where the lower (π,π) spin-density-wave precursor band at the antinodal point crosses ω=0. For U=7 and T=0.0714 the numbers are δ_max=0.15 and crossing at δ=0.16; for U=3.69 they are 0.1 and 0.11. The same spectral crossing produces a maximum in the uniform spin susceptibility χ_sp(0,0)(δ), which survives to higher temperatures than the compressibility maximum because it does not require the Fermi function to resolve the peak. In both interaction regi","pith_inferences":["If the spectral-crossing mechanism is right, the compressibility maximum is a thermodynamic proxy for the antinodal pseudogap edge, so the crossover line δ_max(T) could be mapped with equation-of-state measurements alone, without resolving the spectrum.","The same crossing logic suggests that other thermodynamic quantities sensitive to the Van Hove singularity, such as the doping derivative of entropy or the specific heat coefficient, should show extrema at nearby dopings; this is a testable extension not made in the paper.","Because TPSC+ underestimates spin correlations by 27–33% at U=7, exact numerical methods may place δ_max and the spectral crossing at slightly different dopings; if the two features remain tied, the mechanism survives, but if they separate, the quantitative coincidence is an artifact of the approximation.","The predicted negative effective charge vertex at low temperature implies that density correlations should develop incommensurate structure; a low-temperature measurement of the density structure factor near δ_max could test the precursor-to-stripe scenario."],"forward_implications":["If the central claim is correct, the cold-atom compressibility maximum is a pseudogap-to-Fermi-liquid crossover, not a signature of Mott criticality, and it should appear even at weak interaction strengths.","The predicted maximum in the uniform spin susceptibility χ_sp(0,0)(δ), measurable as a Knight shift, should appear at somewhat higher temperatures than the compressibility maximum and could be easier to observe experimentally.","As temperature decreases, δ_max(T) moves away from half-filling toward the quantum critical point of the spin-density-wave transition.","At sufficiently low temperature, incommensurate spin fluctuations should produce more than two SDW precursor peaks in the spectral function and density of states, a signature accessible to photoemission-type measurements.","The large charge compressibility deep in the pseudogap regime, combined with an effective charge vertex that can turn negative, raises the possibility of SDW-driven charge-density-wave or stripe precursors at very low temperature."],"fun_headline_variants":["Compressibility peak signals pseudogap-to-Fermi-liquid switch","Knight shift maximum predicted at same doping as compressibility peak","Spin-density precursor sets compressibility maximum","Hubbard model: compressibility max locates Fermi liquid onset","Pseudogap crossover identified by compressibility and Knight shift"],"cache_read_input_tokens":37632,"weakest_assumption_plain":"The load-bearing premise is that TPSC+ is quantitatively trustworthy at U≈7 and T≈0.0714, the regime where the paper itself reports that it underestimates spin correlations by 27–33% and underestimates pseudogap effects, so that the predicted value of δ_max and its coincidence with the spectral crossing are not shifted by the approximation.","fun_headline_variants_meta":{"raw":{"variants":["Compressibility peak signals pseudogap-to-Fermi-liquid switch","Knight shift maximum predicted at same doping as compressibility peak","Spin-density precursor sets compressibility maximum","Hubbard model: compressibility max locates Fermi liquid onset","Pseudogap crossover identified by compressibility and Knight shift"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000274,"raw_usage":{"total_tokens":1588,"prompt_tokens":966,"completion_tokens":622,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":710,"completion_tokens_details":{"reasoning_tokens":539}},"tokens_in":710,"tokens_out":622,"duration_ms":38483,"temperature":1.0,"reasoning_tokens":539,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T03:55:51.096654+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A determinant or diagrammatic quantum Monte Carlo calculation at U=7 and T=0.0714 in the thermodynamic limit, locating both the maximum of κ(δ) and the doping at which the antinodal SDW precursor crosses ω=0; if these two dopings differ beyond numerical uncertainty, or if χ_sp(0,0)(δ) has no maximum at T=0.1, the central claim is refuted.","supporting_citations":[],"review_version":1}