{"id":"f3f77aba-6be6-4aff-976f-1a4b5c05fb23","arxiv_id":"2412.16289","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Coadjoint-orbit bosonization shows the de Haas-van Alphen phase shift is governed by the static anomalous Hall conductance, with Berry-curvature corrections to the Lifshitz-Kosevich amplitude.","lead":"This paper develops a field-theoretic description of electrons in a magnetic field using coadjoint-orbit bosonization, and derives the de Haas-van Alphen oscillations from a topological term in the action. It shows that for interacting systems the oscillation phase is set by the static anomalous Hall conductance, and it calculates Berry-curvature corrections to oscillation amplitudes.","discovery_kind":"unification","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Interacting extension in Supplement §IV.A assumes the WZW term is fixed by the kinetic-momentum Fermi-surface area; this step-function replacement is asserted, not proven.","rationale":"The reader's weakest-assumption analysis identifies exactly the same load-bearing point: the interacting proof in Supplement §IV.A assumes the free-fermion WZW term survives with f0 as a step, and that the phase shift is fixed by the kinetic-momentum Fermi-surface area. I agree that this is the most exposed part of the paper. The free-fermion derivation of Eq. (15) is careful and appears internally consistent; the novel claim for interacting systems, Eq. (20), is where the argument relies on an assertion about the universality of the WZW term and the Luttinger relation in a magnetic field. The paper itself flags in the Discussion that amplitude corrections from interactions are deferred, but the phase-shift claim is presented as fully non-perturbative, so the WZW step-function replacement carries the entire weight of the interacting extension. The reader's CONDITIONAL verdict is appropriate: the central claim is plausible and well-supported for free fermions, but the interacting step lacks a rigorous justification. My proposed test would settle the concern by checking, at first order in interactions, whether the dHvA phase extracted from the exact density matrix matches 4π²σ_H. No change to the reader's verdict is needed.","tokens_in":20248,"tokens_out":15831,"duration_ms":149386,"concrete_test":"Re-derive the zero-mode action in Supplement §IV.A without the step-function replacement, keeping the full interacting one-body density matrix f0(π) to first order in the interaction strength, for a 2D lattice model with nonuniform Berry curvature (e.g., the Haldane model with a Hubbard U term). Extract the dHvA phase shift from the O(B^0) coefficient of the θ-term and independently compute the static Hall conductance σ_H via the Streda formula for the same model. If the phase shift differs from 4π²σ_H at O(U), the step-function assumption fails and Eq. (20) is falsified.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim, Eq. (20), that the dHvA phase shift equals 4π²σ_H for generic interacting systems, rests on the derivation in Supplement §IV.A. There, the interacting zero-mode action is obtained by repeating the free-fermion derivation with f0(π,B) taken as a step function (0 or 1 on either side of the Fermi surface), so that the topological θ-term and the Kac-Moody braiding algebra are determined solely by the kinetic-momentum Fermi-surface area A_π_FS(B). This is the load-bearing step. The WZW term is S_WZW = ∫ f0 U⁻¹ i∂t U, and for an interacting ground state f0 is the exact one-body density matrix, which has a quasiparticle weight Z < 1 and a smooth tail, not a sharp step. The O(B^0) coefficient of the θ-term is not shown to reduce to A_π_FS(B) alone; it contains ∫ f0 Ω and possible interaction corrections. Moreover, the proof that A_π_FS(B) satisfies the Luttinger relation ν = A_π_FS/(2π)² in a magnetic field for interacting or non-Fermi-liquid systems is not provided. If interactions renormalize the WZW term or violate Luttinger in a field, the dHvA phase shift will not equal 4π²σ_H.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops a coadjoint-orbit bosonization framework for multi-orbital fermionic systems in a weak magnetic field. The authors project the bosonized action onto a single band, and show that Berry curvature enters through the phase-space measure, the Moyal bracket, and the effective dispersion. The zero-mode sector of the theory yields de Haas-van Alphen oscillations with a Berry-phase-induced phase shift and a modified Lifshitz-Kosevich formula. The central new claim is that for interacting systems the dHvA phase shift is determined by the static anomalous Hall conductance, leading to Eq. (20). The supplemental material provides the band-projection details, the anisotropic-Fermi-surface analysis, the Poisson-resummation derivation of the Lifshitz-Kosevich formula, and the braiding-algebra argument connecting the dHvA phase to sigma_H.","tokens_in":20553,"tokens_out":5830,"duration_ms":54405,"significance":"The free-fermion part of the paper is a solid technical contribution: the star-diagonalization band projection, the appearance of Berry curvature in the phase-space measure and Moyal product, the topological theta-term in the zero-mode action, and the resulting modified Lifshitz-Kosevich formula are carefully derived and checkable. The prediction of an additional temperature-dependent factor in the oscillation amplitude, Eq. (16), is falsifiable, and the paper contains no fitted parameters. However, the advertised generalization to interacting systems, encapsulated in Eq. (20), is the most important new claim and currently rests on an unproven step-function replacement for the exact one-body density matrix. If that step can be supplied, the result would be significant: it would replace the single-particle Berry phase by the static anomalous Hall conductance in dHvA and connect two widely studied observables. As written, the interacting claim needs additional support.","major_comments":[{"comment":"Eq. (S54) writes the WZW term as ∫ f0(π,B) U^{-1} i∂t U and asserts that f0(π,B) equals 0 or 1 on either side of the Fermi surface. For an interacting ground state, f0 is the exact one-body density matrix, which has quasiparticle weight Z < 1 and smooth tails; the O(B^0) coefficient of the resulting theta-term is not shown to reduce to Aπ_FS(B)/2πB. Since this step is the entire basis for extending the dHvA phase shift from Eq. (15) to Eq. (20), the central claim for generic interacting systems is currently an assertion rather than a derivation.","section":"Supplement §IV.A, Eq. (S54)"},{"comment":"The IR braiding calculation uses kF;x,y(θ) and the single-particle Berry connection A(θ), and explicitly invokes kx,y = πx,y ∓ B Ay,x (Eq. (S48)), which is a free-fermion band-projection relation. For a generic interacting system these single-particle objects are not defined, so the evaluation of the braiding algebra in Eqs. (S51)–(S52) cannot be performed as written. The paper needs either a derivation of these IR operators directly from the interacting WZW term or an explicit argument that the same algebra is obtained; otherwise the relation between the dHvA phase and σH is not established beyond free fermions.","section":"Supplement §IV.A, Eqs. (S47)–(S52)"},{"comment":"Eq. (19) relies on the Streda formula σH = dν/dB and on the identification ν = A_FS/(2π)^2 in the presence of a magnetic field. The latter Luttinger-type relation is not proved for interacting or non-Fermi-liquid systems; the reference to Ref. [62] concerns static magnetic response of the density, not the in-field Luttinger theorem needed here. If the in-field Luttinger relation fails, the replacement A_FS/B + γ = A0_FS/B + 4π²σH in Eq. (20) would be invalid.","section":"Main text, Eq. (19)"}],"minor_comments":[{"comment":"The quantities λ1 and λ2 in Eq. (16) are not defined in the main text; please define them or refer explicitly to the supplement where they are introduced.","section":"Main text, Eq. (16)"},{"comment":"The expression A(x) and A(p) is introduced without defining what these quantities are; A(p) is presumably the Berry connection in momentum space, but this should be stated explicitly to avoid confusion with the vector potential A(x).","section":"Main text, after Eq. (5)"},{"comment":"There is a typographical error in the first paragraph of Supplement II: 'natrual' should be 'natural'.","section":"Supplement II"},{"comment":"The figure caption labels the Berry phase but does not indicate how it is visualized in the figure; a sentence clarifying that γ = ∮ A·dk is the Aharonov-Bohm-like phase accumulated along the Fermi surface would improve readability.","section":"Fig. 1"}],"recommendation":"major_revision","confidential_remarks":"The paper is likely of interest to the journal if the interacting claim can be made rigorous. The free-fermion derivation is solid, but the title and abstract promise the interacting result; as written, that result rests on a nontrivial assumption. I would encourage a revision that either proves the step-function replacement from the Kac-Moody algebra and Luttinger theorem, or substantially weakens the claim to free fermions plus a clearly labeled conjecture for interacting systems."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear Colleague,\n\nThe thing you should know: this paper develops a bosonized field theory for multi-orbital fermions and derives Berry phase effects in de Haas-van Alphen (dHvA) oscillations, including a modified Lifshitz-Kosevich formula. The headline result is that for interacting systems, the dHvA phase shift is exactly 4π²σ_H (the static anomalous Hall conductance). The free-fermion derivation is careful; the interacting extension is the soft spot.\n\nWhat's genuinely new: (1) a systematic band projection within coadjoint-orbit bosonization that produces Berry curvature modifications to the phase-space measure and the Moyal product; (2) an explicit Lifshitz-Kosevich formula with Berry curvature corrections to the amplitude (the cyclotron frequency becomes B-dependent); (3) a non-perturbative braiding-algebra argument linking the dHvA phase to σ_H via the Streda formula. The supplement's star-diagonalization is well executed, and it also reproduces the semiclassical Boltzmann equation with the 1+BΩ density-of-states factor. The derivation of the zero-mode action and its mapping to a θ-term is clean.\n\nWhere I'd press: the interacting claim rests on an assumption in Supplement §IV.A that the WZW term uses f0(π,B)—the exact ground-state one-body density matrix—as a step function (0 or 1). For any interacting system, f0 has quasiparticle weight Z<1 and a smooth tail; the θ-term coefficient is ∫ f0 Ω, so it's not automatically the Fermi surface area. The paper asserts, rather than proves, that the non-step part doesn't shift the phase. The Luttinger theorem in a field for non-Fermi liquids is also cited without proof. If that fails, Eq. (20) goes with it. This doesn't make the paper wrong—the result may well be true—but the central claim for interactions is not yet derived.\n\nMinors: the amplitude corrections are for free fermions; interactions are deferred. The paper relies heavily on the authors' prior work [38], which is appropriate here.\n\nWho this is for: condensed matter theorists working on quantum oscillations, Berry phase physics, or nonlinear bosonization. It deserves a serious referee. I'd send it to peer review and ask the authors to justify the step-function f0 in the interacting case.","headline":"Strong free-fermion derivation of Berry phase in dHvA; the interacting extension to σ_H is a bold but unproven step-function assumption.","tokens_in":21029,"tokens_out":5194,"would_cite":true,"duration_ms":43740,"reading_group":"yes","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 argues that the de Haas-van Alphen phase shift in interacting metals is set by the static anomalous Hall conductance, not the single-particle Berry phase.","keywords":["de Haas-van Alphen effect","Berry phase","anomalous Hall conductance","coadjoint orbit bosonization","multiorbital Fermi surface","Lifshitz-Kosevich formula","non-Fermi liquid","orbital magnetization"],"falsifier":"Measure the static anomalous Hall conductance and the de Haas-van Alphen phase shift in the same 2D correlated metal, extracting the phase from the intercept of the oscillations plotted against $1/B$; if the phase differs from $4\\pi^2\\sigma_H$ beyond the combined experimental uncertainty, Eq. (20) is falsified. A microscopic test would be to compute the WZW term at first nontrivial order in interactions and check whether it acquires a correction not fixed by $A^\\pi_{\\mathrm{FS}}(B)$.","tokens_in":20046,"feed_emoji":"🧲","tokens_out":7551,"duration_ms":63538,"temperature":0.7,"pith_summary":"The paper builds a low-energy field theory for a multi-orbital Fermi surface by bosonizing the single-particle distribution function through coadjoint orbits. Projecting this theory onto a single band in a weak magnetic field makes Berry curvature appear in the phase-space measure, the Moyal product, and the effective dispersion. The central result is that for a generic interacting system the de Haas-van Alphen oscillation of orbital magnetization is phase-shifted by an amount fixed by the static anomalous Hall conductance, $4\\pi^2\\sigma_H$, replacing the single-particle Berry phase. The paper also delivers a modified Lifshitz-Kosevich formula in which the amplitude carries Berry-curvature and orbital-magnetic-moment corrections through a field-dependent cyclotron frequency.","feed_headline":"Quantum oscillation phase shift equals anomalous Hall conductance","feed_subtitle":"A bosonization argument shows interacting metals set their dHvA phase by the anomalous Hall conductivity, not just by a Berry phase.","key_machinery":"The load-bearing object is the zero-mode sector of the bosonized action, $S_{\\mathrm{zero}}[p,q]=\\int dt[(-A_{\\mathrm{FS}}/(2\\pi B)-\\gamma/(2\\pi))\\dot q+p\\dot q-\\bar{\\omega}_c p^2/2]$, which describes a particle on a ring threaded by a flux: the coefficient of $\\dot q$ is a topological $\\theta$-term that survives interactions because it descends from the Hamiltonian-independent Wess-Zumino-Witten term. The same WZW term, together with the Kac-Moody algebra of the density modes, lets the translation-operator braiding be evaluated in both the UV and the IR, matching $\\gamma$ to the anomalous Hall conductance via the Streda formula. Berry curvature also enters through the modified phase-space measure $1+B\\Omega$, the Moyal product, and the renormalized cyclotron frequency $\\bar{\\omega}_c$, which is what changes the oscillation amplitudes.","core_discovery":"On its own terms, the paper's central claim is Eq. (20): for a generic interacting 2D metal the oscillatory part of the free energy is $F_{\\mathrm{osc}}=\\sum_k A_k\\cos[k(A^0_{\\mathrm{FS}}/B+4\\pi^2\\sigma_H)]$, so the phase shift in dHvA oscillations exactly matches the static anomalous Hall conductance, $\\sigma_H(q\\to 0,\\omega=0)$. For noninteracting systems this reduces to the familiar Berry-phase shift $\\gamma$, and the paper's non-perturbative braiding-algebra argument shows that for Fermi liquids and non-Fermi liquids the same combination $\\gamma/(4\\pi^2)+\\partial_B A_{\\mathrm{FS}}/(4\\pi^2)$ continues to control both the dHvA phase and the Hall response. The proof runs through the topological Wess-Zumino-Witten term in the bosonized action, which is independent of the Hamiltonian, so only the kinetic-momentum Fermi-surface area and the Kac-Moody braiding algebra are needed.","pith_inferences":["A natural extension, which the paper leaves open, is the 3D version: if the same $\\theta$-term logic carries over, dHvA phase shifts in 3D correlated metals would also be tied to the Hall response, with the Fermi-volume area replacing the 2D Fermi-surface area.","Because the argument is Hamiltonian-independent, a testable prediction is that the dHvA phase shift and the anomalous Hall conductance should track each other under doping or pressure in the same material, even when the quasiparticle description changes sharply.","The amplitude correction $\\exp(\\lambda_2 T)$ suggests an experimental route to separate Berry-curvature contributions from interaction-induced renormalizations: measure the temperature dependence of dHvA amplitudes at fixed field and look for the sign and magnitude of $\\lambda_2$."],"forward_implications":["If Eq. (20) is correct, dHvA experiments in correlated 2D metals can be used to read off the static anomalous Hall conductance from the oscillation phase, without needing a separate transport measurement.","The modified Lifshitz-Kosevich formula predicts an amplitude factor $\\exp(\\lambda_2 T)$ beyond the usual $\\exp(-\\lambda_1 T/B)$, a signature of Berry-curvature and orbital-moment corrections testable in materials with strong Berry curvature on the Fermi surface.","The non-perturbative proof extends the dHvA phase-shift relation to non-Fermi liquids, where quasiparticle Berry phases are not defined, as long as the kinetic-momentum Fermi surface remains well defined.","The cubic $\\phi^3$ corrections produce a small temperature-dependent phase shift for non-parabolic bands, which the paper connects to observations in 3D metals."],"supporting_citations":[{"why":"It supplies the companion coadjoint-orbit bosonization in a weak magnetic field, including the WZW term, mode expansion, and the topological term that produces dHvA.","marker":"[38]"},{"why":"It establishes the coadjoint-orbit nonlinear bosonization method used to incorporate interactions.","marker":"[33]"},{"why":"It provides the conventional Lifshitz-Kosevich formula that this work modifies.","marker":"[19]"},{"why":"It gives the Berry-phase correction to the phase-space density of states and the effective cyclotron frequency used in the amplitudes.","marker":"[25]"},{"why":"It identifies the anomalous Hall effect as a Fermi-surface Berry-curvature property, which the braiding argument builds on.","marker":"[61]"},{"why":"It supplies the prior perturbative interacting result for the static magnetic response that the paper generalizes non-perturbatively.","marker":"[62]"},{"why":"It provides the Kac-Moody density algebra and Luttinger-theorem derivation used for the UV/IR braiding matching.","marker":"[58]"}],"fun_headline_variants":["dHvA phase shift equals anomalous Hall conductance","Hall conductance sets dHvA oscillation phase","Interacting metals: dHvA phase from Hall conductance","Bosonization links Berry phase to Hall response"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The derivation assumes that in the interacting system the low-energy bosonized action keeps the free-fermion Wess-Zumino-Witten term unchanged, so the dHvA phase is set only by the kinetic-momentum Fermi-surface area; if interactions renormalize that term or the braiding algebra beyond what the Fermi-surface area encodes, the phase-shift relation breaks.","fun_headline_variants_meta":{"raw":{"variants":["dHvA phase shift equals anomalous Hall conductance","Hall conductance sets dHvA oscillation phase","Interacting metals: dHvA phase from Hall conductance","Bosonization links Berry phase to Hall response"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000305,"raw_usage":{"total_tokens":1733,"prompt_tokens":913,"completion_tokens":820,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":529,"completion_tokens_details":{"reasoning_tokens":757}},"tokens_in":529,"tokens_out":820,"duration_ms":7649,"temperature":1.0,"reasoning_tokens":757,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T10:44:40.879070+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the static anomalous Hall conductance and the de Haas-van Alphen phase shift in the same 2D correlated metal, extracting the phase from the intercept of the oscillations plotted against $1/B$; if the phase differs from $4\\pi^2\\sigma_H$ beyond the combined experimental uncertainty, Eq. (20) is falsified. A microscopic test would be to compute the WZW term at first nontrivial order in interactions and check whether it acquires a correction not fixed by $A^\\pi_{\\mathrm{FS}}(B)$.","supporting_citations":[{"cited_title":"Lifshitz and A","cited_arxiv_id":null,"evidence_quote":"It provides the conventional Lifshitz-Kosevich formula that this work modifies."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It identifies the anomalous Hall effect as a Fermi-surface Berry-curvature property, which the braiding argument builds on."},{"cited_title":"Chen, Static magnetic response of non-fermi-liquid density, Phys","cited_arxiv_id":null,"evidence_quote":"It supplies the prior perturbative interacting result for the static magnetic response that the paper generalizes non-perturbatively."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It provides the Kac-Moody density algebra and Luttinger-theorem derivation used for the UV/IR braiding matching."}],"review_version":1}