{"id":"563c7732-f034-4d73-8dc1-80871e0a25e4","arxiv_id":"2505.00069","paper_version":2,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"An inhomogeneous superconductor's superfluid weight must include the response of the pairing potential to the vector potential; for unpinned vortex lattices this correction cancels the naive value completely.","lead":"This paper derives a formula for the superfluid weight of superconductors with strongly inhomogeneous pairing, including the response of the pairing potential to the probing vector potential. The correction is negligible for uniform systems but essential for periodic potentials and, especially, for vortex lattices, where the uncorrected formula wrongly predicts a finite superfluid weight.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Vortex-lattice null space of K is unanalyzed; the D_s = 0 cancellation rests on numerical agreement.","rationale":"The reader and I converge on the same weakest point: the one-dimensional-nullspace assumption for K in the vortex-lattice limit. I do not treat this as a demonstrated error, because the paper's numerical validation against d^2F/dA^2 (Fig. 1(b)) and the flat E(Q_x) in SM Fig. 2 directly support the physical outcome D_s = 0. The flat energy means the uncorrected result is indeed corrected to zero by the response term, and the free-energy derivative is an independent route that does not pass through the pseudoinverse. However, the formal derivation has a gap: the rank-one-less claim is not verified for a configuration with near-degenerate translations, and the cancellation is delicate. A concrete SVD check would settle it. Because the evidence currently favors the claim and the issue is a missing analysis rather than a contradiction, I keep the reader's ACCEPT verdict (unchanged) but flag that the authors should supply the null-space spectrum or the robustness risk remains latent.","tokens_in":21886,"tokens_out":16096,"duration_ms":185152,"concrete_test":"Recompute the SVD of K (Eq. 12) for the V0 = 0 vortex-lattice state used in Fig. 1(b), and for at least two larger magnetic supercells, recording the two smallest singular values and the overlaps of the corresponding right singular vectors with (i) the global-phase mode iΔ_m and (ii) the finite-difference response δΔ/δQ_x obtained from self-consistently converged solutions at small Q_x using the smooth-phase convention of the SM. Also check the inner product of C_μ with each near-null right singular vector. If the second-smallest singular value does not tend to zero with system size, or if the pseudoinverse solution differs from the finite-difference δΔ/δQ_x along the translation direction by more than numerical tolerance, the D_s = 0 cancellation is not robust and the central claim for unpinned vortex lattices would need revision.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The flagship result is D_s = 0 for an unpinned vortex lattice, obtained as a cancellation between the uncorrected D_s^(0) and the correction δD_s in Eq. (16). The correction is computed from δΔ/δA obtained via the pseudoinverse of K in Eq. (12). The paper asserts (SM 'Inversion of Equation (12)') that K has exactly one zero mode, the global U(1) phase of Δ, and that zeroing a single singular value gives the physical response. This assertion is not proven for the vortex-lattice case. At V0 = 0 the vortex array has quasi-continuous translational degeneracy, reflected in the flat E(Q_x) in SM Fig. 2; the corresponding order-parameter deformation (a magnetic-translation-covariant spatial derivative of the vortex-lattice solution) is a near-zero mode of the linearized gap equation. If K has an additional zero or near-zero singular value, the standard pseudoinverse either projects out a physical translation response or amplifies it; because D_s = 0 is a complete cancellation, a small error in δΔ/δA would produce a nonzero D_s. The manuscript does not report the singular-value spectrum, the null-space dimension, or the overlap of the source C_μ with near-null-space vectors in the vortex case; the only support is the numerical agreement with d^2F/dA^2 in Fig. 1(b). This is the most load-bearing weakness because the qualitative claim that the uncorrected expression is wrong for vortices rests entirely on this cancellation being robust.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper derives a linear-response expression for the superfluid weight D_s of strongly inhomogeneous superconductors, including the response of the pairing potential Δ to the probe vector potential A. The central formula is D_s = D_s^(0) + δD_s, with δD_s = 2 Re Σ_m (C_μ)_m δΔ_m^*/δA_ν, where δΔ/δA is obtained from the singular linear system K δΔ = C (Eqs. 12–16). The authors argue that the correction is negligible only for homogeneous pairing, significant for periodically modulated pairing, and essential for vortex lattices: for an unpinned vortex lattice the uncorrected D_s^(0) is finite while the full expression gives D_s = 0. The paper validates the formula by comparing with the numerically computed second derivative d²F/dA² of the free energy for both a periodic-potential superconductor and a vortex lattice.","tokens_in":22177,"tokens_out":12083,"duration_ms":132451,"significance":"If the result is correct, it is a practically useful and conceptually important contribution: it provides a concrete BdG-level formula for the superfluid weight in the presence of strong spatial variations of the pairing potential, including the vortex case where collective-mode corrections are qualitatively decisive. The derivation in the main text and Supplementary Material is explicit and self-contained, and the comparison with the independent free-energy second-derivative benchmark is a genuine strength. The gauge-invariance discussion, in particular the projection of the global phase mode, is thoughtful. The claim that the standard uncorrected expression is qualitatively wrong for an unpinned vortex lattice is physically expected from earlier work [66–68], which lends additional credibility to the numerical cancellation. The remaining major weakness is that the null-space structure of the linear-response kernel K in the vortex case is asserted rather than analyzed; this is addressable and does not, in my view, invalidate the central derivation.","major_comments":[{"comment":"The proof that K has exactly one zero mode is incomplete for the vortex-lattice case. In the unpinned case V0=0 the vortex lattice spontaneously breaks continuous translational symmetry, and a uniform translation of the vortex array is a zero-energy Goldstone deformation; the corresponding order-parameter change δΔ_m ∝ v·∇Δ_m is a periodic function on the magnetic supercell and hence lies in the space on which Eq. (12) acts. The manuscript asserts (SM, 'Inversion of Equation (12)') that K has rank one less than its dimension, reflecting only the global U(1) phase, but it provides no singular-value spectrum, no nullity count, and no overlap of C_μ with near-null vectors for the vortex case. This matters because the central vortex result D_s=0 is obtained as the complete cancellation D_s^(0)+δD_s=0 in Fig. 1(b): if the pseudoinverse projects out a physical translational mode, Eq. (16) will not yield the correct δD_s. The agreement with d²F/dA² shown by the yellow crosses is encouraging, but it is a single numerical benchmark, not an analysis of the kernel. The authors should report the singular-value spectrum of K in the vortex case, the dimension of the near-null space, the overlap ⟨C_μ|null⟩, and a system-size convergence study of the cancellation.","section":"SM 'Inversion of Equation (12)'; Fig. 1(b)"},{"comment":"The finite-difference evaluation of δΔ_m/δA_μ for a vortex lattice is delicate because the vortex phase is singular and the numerical solution for Δ(Q) is not guaranteed to be a smooth function of Q. The authors correctly note that δD_s must be invariant under adding a large complex multiple z_i Δ_m, and they state that they verified numerically that Eq. (84) projects out such terms. However, no quantitative description of this verification is given (for example, the range of |z_i| tested or the residual variation of δD_s). Since the vortex-lattice conclusion rests on this projection in the finite-difference route, the paper should either present this check explicitly or base the vortex result on the SVD route together with the null-space analysis requested above.","section":"SM 'Gauge Invariance'; Eq. (93)"}],"minor_comments":[{"comment":"The pairing contribution to the current is written as [δΔ c†c† − (1/2U)δ|Δ|² + H.c.]; because the bracket includes H.c., the total constant term is −(1/U)δ|Δ|², consistent with SM Eq. (31). The present notation invites a spurious factor-of-two concern; please write the total term explicitly as −(1/U)δ|Δ|² to avoid confusion.","section":"Eq. (2) and SM Eq. (31)"},{"comment":"The sentence 'only the value of D_s obtained by taking into account the corrections due to δΔ_m/δA_μ = 0 agrees...' appears to contain a typo; the intended meaning is clearly δΔ_m/δA_μ ≠ 0, since the whole point is that the corrections are nonzero for V0 ≠ 0.","section":"Fig. 1(a) discussion"},{"comment":"SM Fig. 2 refers to 'Fig. 2 of the main text', but the main text contains only Fig. 1; the caption should reference Fig. 1(b).","section":"SM Fig. 2 caption"},{"comment":"Reference [59] has a formatting error: 'Phys. Rev. B 89, 014507 (2014)89, 014507 (2013)' should be cleaned up.","section":"Reference [59]"},{"comment":"After Eq. (17), the text says 'for V0=0 the superconductor is homogeneous'; this is only true for the periodic-potential panel. In the vortex-lattice panel V0=0 still leaves the vortex-induced inhomogeneity. Consider clarifying to avoid ambiguity.","section":"Main text around Eq. (17)"}],"recommendation":"major_revision","confidential_remarks":"The main risk is the unanalyzed null-space structure of K in the vortex-lattice case; this is load-bearing because the D_s=0 result is a complete cancellation. The rest of the derivation is coherent and the numerical benchmark is strong, so I would be comfortable with acceptance after the authors supply the singular-value spectrum, nullity, overlap analysis, and a finite-size scaling check for the vortex case."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear Colleague,\n\nTwo things to know. First, the central result of this paper is correct and genuinely useful: computing the superfluid weight of a strongly inhomogeneous superconductor requires a correction from the response of the pairing potential to the vector potential, and the authors give the first explicit real-space linear-response equation for that response. Second, the qualitative claim—that the uncorrected expression gives a finite superfluid weight for an unpinned vortex lattice when the true value is zero—is supported by numerical comparison with the free-energy second derivative, but the paper does not analyze the null space of the linear-response kernel in the vortex case. That leaves a small, real gap in the argument.\n\nWhat is new: vertex corrections and collective-mode contributions are known from Refs. 20, 26, 47-65, and the authors say so. The new piece is the explicit equation K δΔ = C for δΔ/δA in a BdG system with strong spatial inhomogeneity, plus the demonstration that including this correction fixes the quantitative error for periodic potentials and the qualitative error for vortices. The derivation is coherent, and the validation against d²F/dA² is a proper independent check. The unpinned vortex-lattice result D_s = 0 is a nice application, even though it is expected from earlier work [66-68].\n\nSoft spots, in proportion. The stress-test about the vortex-lattice null space is the one that matters. In the SM the authors assert that K has rank one less than its dimension and prove that the global U(1) phase is a zero mode. They do not prove that there are no other zero or near-zero modes arising from the quasi-continuous translational degeneracy of the vortex array. If such a mode couples to the source C, a naive pseudoinverse could suppress or amplify the physical response, and the D_s = 0 cancellation would be fragile. The numerical agreement with the free-energy derivative is reassuring, but it is a single benchmark. I would ask the authors to report the singular-value spectrum of K for the vortex case, or at least the overlap of C with the near-null-space vector, before publication. That is a revision-level request, not a rejection.\n\nMinor issues: a factor-of-two mismatch in the |Δ|²/U term between Eq. (1)/(2) and SM Eq. (31), and no code or data shipped. The citation pattern is appropriate, including the self-citation to the authors' own prior implementation.\n\nWho this is for: anyone computing superfluid weight in moiré flat bands, disordered films, or vortex states. It deserves a serious referee. My recommendation: send it out, with a request for the null-space analysis. The central derivation holds up; the gap is a robustness check, not a fundamental flaw.","headline":"A clear and useful derivation of the pairing-potential response correction to superfluid weight, with a solid numerical benchmark but a genuine gap on the vortex-lattice null space.","tokens_in":22682,"tokens_out":5090,"would_cite":true,"duration_ms":52057,"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":"The paper derives a linear-response correction to the superfluid weight from the pairing potential's response to a vector potential and shows it is essential for inhomogeneous superconductors, including setting the unpinned vortex-lattice…","keywords":["superfluid weight","superfluid stiffness","inhomogeneous superconductivity","pairing potential response","linear response theory","vortex lattice","Bogoliubov-de Gennes","Berezinskii-Kosterlitz-Thouless transition"],"falsifier":"Compute the null space of $K$ (Eq. 12) for a vortex-lattice configuration with no pinning: if the dimension exceeds one, the pseudoinverse projects out a physical mode and the predicted $D^{(s)}_{\\mu\\nu}=0$ would not survive under a different gauge choice or finite-size extrapolation. Alternatively, evaluate $D^{(s)}$ by the finite-difference free-energy method at larger system sizes and check whether the cancellation persists as the vortex lattice is translated by a small amount.","tokens_in":21659,"feed_emoji":"🧲","tokens_out":9683,"duration_ms":87743,"temperature":0.7,"pith_summary":"This paper establishes a linear-response formula for the superfluid weight $D^{(s)}_{\\mu\\nu}$ of a superconductor whose pairing potential $\\Delta$ varies strongly in space. The central claim is that the standard expression, which treats $\\Delta$ as fixed when a vector potential $A$ is applied, misses a correction $\\delta D^{(s)}_{\\mu\\nu}$ coming from the way $\\Delta$ itself responds to $A$. For a periodic modulation of the pairing, the correction is quantitatively necessary; for an unpinned vortex lattice it is qualitative, turning the uncorrected finite stiffness into the expected value $D^{(s)}_{\\mu\\nu}=0$. The paper verifies the formula by comparing it with direct numerical second derivatives of the free energy with respect to $A$. Because the superfluid weight sets the Berezinskii-Kosterlitz-Thouless temperature in two dimensions, the correction matters for predictions of phase coherence in inhomogeneous thin films.","feed_headline":"Vortex lattices lose superfluid weight once gap response is counted","feed_subtitle":"The usual formula skips how the gap shifts under a probe; with vortices it wrongly predicts a finite supercurrent.","key_machinery":"The central object is the pairing-potential response $\\delta\\Delta_m/\\delta A_\\mu|_0$, the change each site's complex gap undergoes when a slow static vector potential is applied. It is computed from the singular linear system $K\\,\\delta\\Delta=C$ (Eqs. 12-15), where the kernel $K$ is the pairing susceptibility including the $-1/U$ term and $C_\\mu$ couples the kinetic-current vertex to pair creation; inverting via a pseudoinverse from a singular-value decomposition projects out the global phase mode of $\\Delta$. This response enters the current operator as off-diagonal particle-hole blocks and produces the correction $\\delta D^{(s)}_{\\mu\\nu}=2\\,\\mathrm{Re}\\sum_m (C_\\mu)_m\\,\\delta\\Delta_m^*/\\delta A_\\nu$. The paper also shows the same result follows from vertex corrections and Ward identities, which is what guarantees gauge invariance of the final tensor.","core_discovery":"Within Bogoliubov-de Gennes mean-field theory for an $s$-wave superconductor with arbitrary spatial inhomogeneity, the paper derives $D^{(s)}_{\\mu\\nu}=D^{(s,0)}_{\\mu\\nu}+\\delta D^{(s)}_{\\mu\\nu}$, where $D^{(s,0)}$ is the usual current-correlation expression and $\\delta D^{(s)}_{\\mu\\nu}=2\\,\\mathrm{Re}\\sum_m (C_\\mu)_m\\, \\delta\\Delta_m^*/\\delta A_\\nu$. The vector $\\delta\\Delta/\\delta A$ is not an independent input: it solves the singular linear system $K\\,\\delta\\Delta=C$ (real and imaginary parts coupled), with $K$ having one zero mode corresponding to the global $U(1)$ phase of $\\Delta$, so the physical solution is obtained by pseudoinverse. The paper argues that this correction is the collective-mode, or anomalous vertex, contribution required by gauge invariance when inhomogeneities mix longitudinal and transverse responses, and demonstrates numerically for two two-dimensional examples that the full formula coincides with the free-energy second derivative while the uncorrected formula does not. In the vortex-lattice example the full formula gives $D^{(s)}_{\\mu\\nu}=0$ for an unpinned array, matching known results, whereas the uncorrected expression is finite.","pith_inferences":["A testable extension the paper does not run: use the corrected stiffness to predict $T_{\\rm BKT}$ in inhomogeneous films and compare with measured sheet inductance or resistance peaks, isolating the $\\Delta$-response contribution from the usual kinetic term.","The null-space assumption could be probed by adding weak disorder that breaks translational invariance; if additional near-zero modes of $K$ appear beyond the global phase, the exact cancellation for vortex arrays may become finite-size dependent.","The same linear-response machinery applies to flat-band superconductors where quantum-geometric contributions are large; the $\\Delta$-response term could partially cancel or enhance those contributions, and the paper's formula provides a way to test which.","The direction dependence of the corrected tensor under an anisotropic pinning potential could serve as a probe of vortex-lattice orientation, since the unpinned cancellation is exact but any pinning symmetry breaking leaves a directional stiffness."],"forward_implications":["For any strongly inhomogeneous superconductor, the superfluid weight requires solving the gap's response to the probe field; the standard uncorrected formula is quantitatively wrong when the pairing modulation is strong.","In two dimensions, because the Berezinskii-Kosterlitz-Thouless temperature is set by the superfluid weight, the correction shifts predicted $T_{\\rm BKT}$ values for patterned and moiré superconductors.","An unpinned vortex lattice has exactly zero superfluid weight, so it produces no Meissner response; the uncorrected expression would falsely predict a finite response.","A periodic pinning potential restores a finite superfluid weight, and the corrected formula interpolates between zero and the uncorrected value as pinning strengthens.","Gauge invariance demands the correction: multiplying the pairing potential by a complex constant produces no total current, and the formula projects out such gauge artifacts."],"supporting_citations":[{"why":"It defines the superfluid weight as the current response to a transverse vector potential, the quantity the paper computes.","marker":"[1]"},{"why":"It gives the standard expression for the superfluid weight that the paper corrects by adding the pairing-potential response.","marker":"[2]"},{"why":"It is an earlier derivation emphasizing the role of the pairing-potential response for orbital-position independence, which the present work extends to rapidly varying phases such as vortex lattices.","marker":"[20]"},{"why":"It identifies the difficulty posed by inhomogeneities that mix longitudinal and transverse responses, motivating the collective-mode correction.","marker":"[45]"},{"why":"It supplies the vertex-correction and Ward-identity relations that justify treating the gap response as the anomalous vertex contribution.","marker":"[47]"},{"why":"It introduces the particle-hole formulation of the superconducting mean-field Hamiltonian used in the linear response calculation.","marker":"[51]"},{"why":"It establishes the vanishing superfluid weight of an unpinned vortex lattice, the result the corrected formula recovers.","marker":"[66]"},{"why":"It confirms the unpinned vortex-lattice result and provides the benchmark against which the uncorrected expression is shown to be qualitatively wrong.","marker":"[67]"},{"why":"It provides the magnetic translation-group description used to place the vortex lattice on a tight-binding lattice.","marker":"[71]"}],"fun_headline_variants":["Gap response zeros superfluid weight for vortex lattices","Superfluid weight formula needs gap response in vortices","Uncorrected superfluid weight flops for vortex lattices","Gap response vital for superfluid weight in inhomogeneous systems","Vortex superfluid weight: include gap response or fail"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the linear-response kernel $K$ has exactly one zero mode, the global phase of the pairing potential, so that the pseudoinverse yields the physical response $\\delta\\Delta/\\delta A$; for a vortex lattice, the vortices' ability to slide without energy cost could create additional zero or near-zero modes, and the paper checks this case by numerical agreement with the free-energy derivative instead of by analyzing the null space.","fun_headline_variants_meta":{"raw":{"variants":["Gap response zeros superfluid weight for vortex lattices","Superfluid weight formula needs gap response in vortices","Uncorrected superfluid weight flops for vortex lattices","Gap response vital for superfluid weight in inhomogeneous systems","Vortex superfluid weight: include gap response or fail"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000924,"raw_usage":{"total_tokens":3979,"prompt_tokens":979,"completion_tokens":3000,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":595,"completion_tokens_details":{"reasoning_tokens":2918}},"tokens_in":595,"tokens_out":3000,"duration_ms":20901,"temperature":1.0,"reasoning_tokens":2918,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T04:54:12.100310+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the null space of $K$ (Eq. 12) for a vortex-lattice configuration with no pinning: if the dimension exceeds one, the pseudoinverse projects out a physical mode and the predicted $D^{(s)}_{\\mu\\nu}=0$ would not survive under a different gauge choice or finite-size extrapolation. Alternatively, evaluate $D^{(s)}$ by the finite-difference free-energy method at larger system sizes and check whether the cancellation persists as the vortex lattice is translated by a small amount.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It defines the superfluid weight as the current response to a transverse vector potential, the quantity the paper computes."},{"cited_title":"Huhtinen, J","cited_arxiv_id":null,"evidence_quote":"It is an earlier derivation emphasizing the role of the pairing-potential response for orbital-position independence, which the present work extends to rapidly varying phases such as vortex lattices."},{"cited_title":"Seibold, L","cited_arxiv_id":null,"evidence_quote":"It identifies the difficulty posed by inhomogeneities that mix longitudinal and transverse responses, motivating the collective-mode correction."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It supplies the vertex-correction and Ward-identity relations that justify treating the gap response as the anomalous vertex contribution."},{"cited_title":"Nambu, Physical Review117, 648 (1960)","cited_arxiv_id":null,"evidence_quote":"It introduces the particle-hole formulation of the superconducting mean-field Hamiltonian used in the linear response calculation."},{"cited_title":"Teitel and C","cited_arxiv_id":null,"evidence_quote":"It establishes the vanishing superfluid weight of an unpinned vortex lattice, the result the corrected formula recovers."},{"cited_title":"Franz and S","cited_arxiv_id":null,"evidence_quote":"It confirms the unpinned vortex-lattice result and provides the benchmark against which the uncorrected expression is shown to be qualitatively wrong."},{"cited_title":"forward hopping","cited_arxiv_id":null,"evidence_quote":"It provides the magnetic translation-group description used to place the vortex lattice on a tight-binding lattice."}],"review_version":1}